Department of College of Engineering The University of Iowa Iowa City, Iowa Loss of Prestress, Camber, and Deflection of Noncomposite and Composite Structures Using Different Weight Concretes Final Report by D. E. Branson B.L.Meyers K. M. Kripanarayanan Report No. 70-6 Prepared Under Iowa State Highway Commission Research Project HR-137 August 1970
254
Embed
Loss of Prestress, Camber, and Deflection of Non-Composite ...publications.iowa.gov/21548/1/IADOT_HR_137_Loss... · of Noncomposite and Composite Structures Using Different ... cretes
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Department of
College of Engineering
The University of Iowa
Iowa City, Iowa
Loss of Prestress, Camber, and Deflection
of Noncomposite and Composite Structures
Using Different Weight Concretes
Final Report
by
D. E. Branson
B.L.Meyers
K. M. Kripanarayanan
Report No. 70-6
Prepared Under Iowa State Highway Commission Research Project HR-137
August 1970
LOSS OF PRESTRESS, CAMBER, AND DEFLECTION
OF NON-COMPOSITE AND COMPOSITE STRUCTURES
USING DIFFERENT WEIGHT CONCRETES
Final Report
by
D, E. Branson Professor of Civil Engineering
B. L. Meyers Associate Professor of Civil Engineering
K. M. Kripanarayanan Research Associate in Civil Engineering
The opinions, findings, and conclusions expressed in this publication are those of the al\thors and not necessarily those of the Iowa State Highway Commission.
Report No. 70-6 Prepared Under Iowa State Highway
Commission Research Project HR-137
Department of Civil Engineering University of Iowa
Iowa City
August 1970
FOREWORD
This is a report of research conducted under the Iowa State
Highway Commission Research Project No. HR-137. The project was
initiated in February 1968. A progress report, No. 69-1, was sub
mitted in February 1969.
This project is being coordinated with the Iowa State Highway
Commission Research Project No. HR-136, "Creep and Shrinkage
Properties of Lightweight Concrete Used in the State of Iowa" (see
final report dated August 1970); and with the Iowa Highway Research
Board Project No. HR-104, ''Field Observation of Five Lightweight
Aggregate Pretensioned Prestressed Concrete Bridge Beams" (see
final report by J. A. Young).
Acknowledgment is made of the assistance of Messrs. S. E.
Roberts, Research Engineer, C. Pestotnik, Bridge Engineer,
Y. H. Gee, Assistant Bridge Engineer, and J. A. Young, Research
Technician, of the Iowa Highway Commission; and Mr. J. H. Boehm
ler, Jr., President, Pres tressed Concrete of Iowa, Inc.
The authors would also like to thank the Idealite Co., Denver,
Colorado, and the Hydraulic Press Brick Co., Brooklyn, Indiana for
donating materials used in the experimental program.
ii
ABSTRACT
Presented in this report are the results of an investigation of
the use of lightweight concretes in prestressed and reinforced con
crete structures. Both "sand-lightweight" and "all-lightweight" con
cretes are included in the study. The sand-lightweight concrete
cons is ts of 100% sand subs ti tu ti on for fines, along with Idea lite coarse
and medium lightweight aggregate and Type I cement. The all-light-
weight concrete consists of Haydite coarse, medium, and fine
aggregates along with Type I cement.
The study is divided into three parts: a materials study of
the concretes themselves, a laboratory study of the behavior of both
non-composite and composite beams that included prestressed (15
beams) and reinforced (3 beams) beams, and the field measurement
of camber of prestressed girders (5 girders) used in the fabrication
of a composite bridge in Iowa. The minimum test period for the
laboratory beams is 6 months, although data is recorded for 1 year
for 3 of the beams. The test period for the bridge girders is 560
days.
The laboratory prestressed concrete beams are designed in
five groups (3 beams in each group) to investigate the loss of
iii
pres tress, initial and time -dependent camber, load -deflection behavior
(under single and repeated load cycles) and the effect of different slab
casting schedules. One group of 3 reinforced beams is used to inves ti -
gate the initial and time-dependent deflection, load-deflection behavior
after sustained loading, and the effect of different slab casting sched-
ules.
The methods described for predicting material behavior and
structural response are generalized to apply to prestressed and rein-
forced structures of normal weight, sand-lightweight, and all-light-
weight concrete. Continuous time functions are provided for all needed
parameters, so that the general equations readily lend themselves to
computer solution. Approximate equations are also included,
Design procedures are presented for the following:
1. Calculation of strength and elastic properties, creep and shrinkage of the lightweight concretes of this project at any time, including ultimate values. An indication is also given of the calcula<· tion of these properties for other concretes in general.
2. Calculation of loss of prestress and camber at any time, including ultimate values, of non-composite and composite prestressed structures.
3. Calculation of deflections at any time, including ulti-mate values, of non-composite and composite reinforced structures.
4. Calculation of deflections of prestressed concrete mem-bers under single and repeated load cycles (with constant as well as increasing stress range). Calculation of deflections of reinforced concrete members under sustained loads in the non-linear range for short times (24 hours) is also included.
iv
Results computed by these methods are shown to be in good
agreement with the control specimen data, the laboratory beam data,
and the bridge girder data.
Published experimental data concerning the time-dependent
(prestress loss, camber, and deflection) effects and load deflection
response of prestressed and reinforced beams are shown to be in
reasonable agreement with the results computed by the design methods
presented in this report. Ranges of variation are also shown. These
data include normal weight, sand-lightweight and all-lightweight con-
crete, non-composite and composite members, and both laboratory
specimens and actual structures.
This project is thought to be the first such comprehensive study
of the initial plus time-dependent material behavior and related struc-
tural response of both non-composite and composite structures using
different weight concretes. A new procedure is also developed for pre-
dieting the entire load-deflection curve of both reinforced and prestressed
members under repeated load cycles into the cracking range.
Keywords: all-lightweight concrete; beams (structural); bridge girders; camber; composite construction (concrete to concrete); creep (materials); deflection; lightweight concrete; loss of prestress; modulus of elasticity; normal weight concrete; precast concrete; prestressed concrete, repeated cycle; sand-lightweight concrete; shrinkage; single cycle; steel relaxation; strain; stress; structural design; sustained; test beams; time-dependent.
v
TABLE OF CONTENTS
Chapter
1.
2.
3.
4.
List of Tables
List of Figures
Notation
INTRODUCTION -
1. 1 Statement of the Problem
1. 2 Objectives and Scope
1. 3 Review of Literature
DESCRIPTION OF EXPERIMENTAL INVESTIGATION
2. 1 Materials and Test Specimens
2.2 Instrumentation and Test Data
STRENGTH AND ELASTIC PROPER TIES, CREEP AND SHRINKAGE
3. 1 Strength and Elastic Properties
3.2 Creep and Shrinkage
LOSS OF PRESTRESS AND CAMBER
4. 1 Relaxation Tests
4. 2 Computed Loss of Prestress, Camber and Deflection
4.3 Required Calculations and Summary of General Parameters
vi
Page
ix
xi
xvii
1
1
2
3
8
8
10
12
12
16
21
21
23
38
TABLE OF CONTENTS (Cont'd)
Chapter
4.4 Sample Calculations
4.5 Experimental Loss of Prestress, Camber and Deflection Results
4.6 Discussion of Experimental Results and Conclusions
4.7 Comparison of Computed and Measured Data Reported by Others
4. 8 Summary of Results Reported by Others and Conclusions
5. LOAD DEFLECTION STUDIES OF PRESTRESSED AND REINFORCED CONCRETE BEAMS
6.
5. 1
5.2
5. 3
5.4
5.5
5.6
General
Single Cycle Load Tests of Prestressed Members
Repeated Load Tests of Prestressed Members
Increasing Load Plus 24-Hour Sustained Load Tests
Results Reported by Others
Summary and Conclusions
SUMMARY AND CONCLUSIONS
LIST OF REFERENCES
APPENDIX A Specimen Details
APPENDIX B Creep and Shrinkage Variables
APPENDIX C Specimen Details for the Data in the Literature
vii
Page
41
44
62
70
89
92
92
93
107
127
132
148
155
161
Ap 1
Ap 12
Ap 20
TABLE OF CONTENTS (Cont'd)
Chapter
APPENDIX D Camber Equations for Common Prestress Profiles
APPENDIX E Photographs of Laboratory Specimens
APPENDIX F Computer Flow Charts and Typical Outputs
viii
Page
Ap 29
Ap 31
Ap 37
LIST OF TABLES
Table Page
1 EXPERIMENTAL AND COMPUTED LOSS OF PRE STRESS FOR LABORATORY BEAMS AND COMPUTED LOSS OF PRESTRESS FOR BRIDGE GIRDERS 54
2 MEASURED AND COMPUTED MIDSPAN CAMBER & DEFLECTION FOR LABORATORY BEAMS & BRIDGE GIRDERS 56
3 COMPUTED ULTIMATE LOSS OF PRESTRESS AT MIDSPAN, BY TERMS, FOR THE LABORATORY BEAMS AND BRIDGE GIRDERS, USING THE GENER -AL EQUATIONS (14) & (17) WITH EXPERIMENTAL PARAMETERS 58
4 COMPUTED ULTIMATE MIDSPAN CAMBER, BY TERMS, FOR THE LABORATORY BEAMS AND BRIDGE GIRDERS, USING THE GENERAL EQS. (15), (16), (18) & (20) WITH EXPERIMENTAL PARAMETERS 60
5 WORKING LOAD, COMPUTED AND OBSERVED VALUES OF ULTIMATE LOAD AS WELL AS VALUES OF WORST DISCREPANCY BETWEEN COMPUTED AND OBSERVED DEFLECTION CURVES 103
6 DETAILS OF REPEATED LOAD CYCLES AND DIS-CREPANCY IN THE OBSERVED AND COMPUTED VALUES OF MIDSPAN DEFLECTION FOR BEAMS OF GRPS C & E 124
7 DETAILS OF INCREASING LOAD PLUS 24-HR SUS-TAINED LOAD TESTS WITH REGARD TO WORKING LOADS, ULTIMATE LOADS AND DEFLECTIONS UNDER THESE LOADS 131
lX
LIST OF TABLES (Cont'd)
Table Page
Al DETAILS OF LABORATORY BEAMS (GRPS A, B, C) AND BRIDGE GIRDERS Ap 2
A2 DETAILS OF LABORATORY BEAMS (GRPS D, E AND F) Ap 4
A3 DETAILS OF CONCRETE MIXES AND MIXING PROCEDURE FOR LT-WT CONCRETES Ap 6
A4 CONCRETE PROPERTIES (GRPS A, B, C AND BRIDGE GIRDERS), TEMPERATURE AND HUMIDITY DATA Ap 7
A5 CONCRETE PROPER TIES (GRPS D, E, & F), TEMPERATURE AND HUMIDITY DATA Ap 9
Ab CONCRETE PROPERTIES OF LAB BEAMS AT "LOAD-DEF" STUDIES Ap 11
Cl PROPERTIES OF TEST BEAMS AT UNIVERSITY OF FLORIDA (~) Ap 21
CZ PROPERTIES OF TEST BEAMS AT UNIVERSITY OF ILLINOIS (24} Ap 22
C3 PROPERTIES OF. TEST BEAMS AT TEXAS A &M UNIVERSITY (27) Ap 23
C4 PROPERTIES OF TEST BEAMS AT UNIVERSITY OF MISSOURI (2_!} Ap 24
C5 DETAILS OF BEAMS REPORTED BY ABELES (2.i) Ap 25
Cb DETAILS OF BEAMS REPORTED BY WARAWARUK, SOZEN & SIESS (Q) Ap 26
C7 DETAILS OF BEAMS REPORTED BY SHAIKH AND BRANSON (49) Ap 27
cs DETAILS OF BEAMS REPORTED BY BURNS & SIESS (2!) Ap 28
x
Figure
1
2
3
4
5
6
7
8
9
10
11
12
LIST OF FIGURES
Laboratory beams and bridge girders
Concrete strength vs time curves for lab concretes (Gps B, C)
Creep coefficient vs time curves for lab concretes (Gps A, B, C)
Shrinkage vs time curves for lab concretes (Gps A, B, C)
Concrete strength vs time curves for lab concrete (Gps E, F)
Creep coefficient vs time curves for lab concrete (Gps D, E, F)
Shrinkage vs time curves for lab concretes (Gps D, E, F)
Results of steel relaxation tests
Determination of experimental loss of prestress
Computed and experimental loss of prestress of beams of Group A (three non-composite beams)
Computed and experimental loss of prestress of beams of Groups B and C (two non-composite and four composite beams)
Computed and experimental loss of prestress of beams of Groups D and E (two non-composite and four composite beams)
xi
Page
9
14
14
14
15
15
15
22
22
45
46
47
Figure
13
14
15
16
17
18
19
20
21
22
23
24
LIST OF FIGURES (Cont'd)
Computed loss of prestress of five composite bridge girders
Computed and experimental midspan camber of beams of Group A (three non-composite beams)
Computed and experimental midspan camber of beams of Groups Band C (two non-composite and four composite beams)
Computed and experimental midspan camber of beams of Groups D and E (two non-composite and four composite beams)
Computed and experimental midspan deflection of beams of Group F (one non-composite and two composite beams)
Computed and experimental midspan camber of five composite bridge girders
Computed and experimental loss of prestress at end of beams reported in Reference (23)
Computed and experimental loss of pres tress at center of beams reported in Reference (~)
Computed and experimental midspan camber of beams reported in Reference (~)
Computed and experimental loss of prestress at center of beams reported in Reference (24)
Computed and experimental midspan camber of beams reported in Reference (24)
Computed and experimental loss of prestress at end of beams reported in Reference (2 7)
xii
Page
48
49
50
51
52
53
72
73
74
77
78
81
Figure
25
26
27
28
29
30
31
32
33
34
35
36
37
LIST OF FIGURES (Cont'd)
Computed and experimental loss of prestress at center of beams reported in Reference (2 7)
Computed and experimental midspan camber of beams reported in Reference (27)
Computed and experimental loss of prestress at end of beam reported in Reference (1.!_)
Computed and experimental loss of prestress at center of beam reported in Reference (31)
Computed and experimental midspan camber of beam reported in Reference (1.!_)
Two point loading for 'load-deflection' studies of laboratory beams
Moment of inertia of cracked section (Icr)
Observed and computed midspan deflection versus load curves for beams of Group A (three non-composite prestressed beams)
Observed and computed midspan deflection versus load curves for beams of Group B (one non-composite and two composite prestressed beams)
Observed and computed midspan deflection versus load curves for beams of Group D (three non-composite prestressed beams)
Details of deflections under repeated loadings
Sample calculations
Observed and computed midspan deflection vs load curve of beam Cl under 3 cycles of repeated loading {one non-composite prestressed beam)
xiii
Page
82
83
86
87
88
94
98
100
101
102
110
114
118
Figure
38
39
40
41
42
LIST OF FIGURES (Cont'd)
Observed and computed midspan deflection versus load curve of beam CZ under 3 cycles of repeated loading (one non-composite prestressed beam)
Observed and computed midspan deflection versus load curve of beam C3 under 3 cycles of repeated loading (one composite prestressed beam)
Observed and computed midspan deflection versus load curve of beam E 1 under 3 cycles of repeated loading (one non-composite prestressed beam)
Observed and computed midspan deflection versus load curves for beams E2 and E3 under 3 cycles of repeated loading (two composite prestressed beams)
Effect of repeated loading (in the cracked range) on total deflections of laboratory beams of Groups C and E
43 Observed and computed values of midspan deflection for beams of Group F under 24-hr sustained loading (one non-composite and two composite reinforced beams)
44
45
46
Observed and computed midspan deflection (using Eqs, (38) and (41) for beams under static loading as in (A) (Data from Reference 56) and as in (B) (Data from Reference 41)
Observed and computed midspan deflection (using Eqs, (38), (40), and (41) for beams under static loading as in (A) (Data from Reference 49) and for beams under repeated loading as in (B) (Data from Reference 54)
Comparison of computed and observed values of midspan deflection of beams in Reference (54), under two cycles of repeated loading (three non-composite reinforced beams)
xiv
Page
119
120
121
122
123
130
134
138
141
LIST OF FIGURES (Cont'd)
Figure Page
47 Range of validity of Eqs. (38), (40) and (41) for rectangular beams with different steel percentages - -included in this dimensionless plot are also the results from studies made on rectangular prestressed beams from Reference (!!_), (49), (54) and (56) as well as the current study 143
48
Bl
B2
B3
Range of validity of Eqs. (38), (40), and (41) for T beams with different steel percentages --included in this dimensionless plot are the results from the current study only
Strain components
Time -dependent strain variables
Nominal creep and shrinkage correction factors for the parameters shown from Ref. 18
E 1 View of laboratory showing beams in foreground and pre-
144
Ap 14
Ap 16
Ap 18
stressing bed containing additional beams at right Ap 32
E2 Forms for beams in prestressing bed Ap 32
E3 Strain gage indicator and switching and balancing unit used with load cells to measure pres tress force
E4 Prestressing bed, jacking equipment and beams stored
Ap 33
in bed Ap 33
ES Close-up of jacking equipment, bulkheads, and grips Ap 34
E6 Shrinkage specimens in foreground and 7 beams ( 1 beam crosswise in foreground). Two additional beams in prestressing bed Ap 34
E7 Two of 4 composite beams. Strain gage points and dial gages can be seen. Strands used in relaxation tests are seen at right Ap 35
xv
LIST OF FIGURES (Cont'd)
Figure
E8 Cylinders loaded in creep racks and Whittemore gage used to measure strains of beams and shrinkage and creep specimens
E9 View of beam C 1 showing the crack pattern prior to failure
ElO View of beam Cl after failure
xvi
Page
Ap 35
Ap 36
Ap 36
NOTATION
l = subscript denoting cast-in-place slab of composite beam or effect of slab
2 = subscript denoting precast beam
A = area of section
A = area of gross section, neglecting the steel g
As = area of tension steel in reinforced members and area of prestressed steel in prestressed members
I
As = area of compression steel in reinforced members and area of non-tensioned steel in prestressed members
At = area of transformed section
a = distance from end of beam to the nearest of 2 symmetrical disphrams. Also used as the distance from end to harped pt. in 2-pt. harping. Also used as empirical constant-see Eq. (1). Also used as distance of load from the near support--see Eq. (41).
b = empirical constant determined in the laboratory--see Eq. (1). Also used as distance between applied loads--see Eq. (41). Also used as compression flange width.
Cs = creep coefficient defined as ratio of creep strain to initial strain at slab casting.
= creep coefficient at any time t
= creep coefficient of the composite beam under slab dead load
= creep coefficient of the precast beam concrete
xvii
c u
C. F.
c
= ultimate creep coefficient defined as ratio of ultimate creep strain to initial strain
= correction factor to account for conditions other than standard
= subscript denoting composite section. Also used to denote concrete, as Ee
cp = subscript denoting creep
D = differential shrinkage strain. Also used as a subscript to denote dead load
DS = subscript denoting differential shrinkage
d = effective depth of section
E = modulus of elasticity
E = modulus of elasticity of concrete such as at 28 days c
E . = modulus of elasticity of concrete at the time of initial Cl
loading, such as at transfer of prestress, etc.
Ecs = modulus of elasticity of concrete at the time of slab casting
Es = modulus of elasticity of steel
e = eccentricity of steel
ec = eccentricity of steel at center of beam. Also used, as indicated, to denote eccentricity of steel in composite section
e0
= eccentricity of steel at end of beam
F = prestress force after losses
Fi = initial tensioning force
F0
= prestress force at transfer (after elastic loss)
xviii
f:,F
f:,F s
= loss of prestress due to time-dependent effects only (such as creep, shrinkage, steel relaxation). The elastic loss is deducted from the tensioning force, Fi' to obtain F
0
= total loss of prestres s at slab casting minus the initial elastic loss that occurred at the time of prestressing
= total loss of prestress at any time minus the initial elastic loss
= total ultimate loss of prestress minus the initial elastic loss
fc = concrete stress at steel c. g. s due to pres tress and pre-cast beam dead load
fed = concrete stress at steel c. g. s due to differential shrinkage
fcs = concrete stress at steel c. g. s due to slab dead load (plus
f ' Sl
f y
H
diaphragm dead load where applicable)
= compressive strength of concrete
= compressive strength of concrete at time t
= compressive strength of concrete at 28 days
= ultimate (in time) compressive strength of concrete
= modulus of rupture of concrete
= tensile strength of concrete
= stress in prestressing steel at transfer (after elastic loss)
= initial or tensioning stress in prestressing steel
= yield strength of steel (defined herein as O. 1% offset)
= relative humidity in percent
= moment of inertia of slab
= moment of inertia of precast beam
xix
= moment of inertia of composite section with transformed slab. The slab is transformed into equivalent precast beam concrete by dividing the slab width by E /E
C2 c 1
= moment of inertia of cracked transformed section
= effective moment of inertia
= moment of inertia of gross section, neglecting the steel
= effective moment of inertia under repeated loads
= moment of inertia of transformed section, such as an uncracked pres tressed concrete section
i ::: subscript denoting initial value
K = deflection coefficient. For example, for beams of uniform section and uniformly loaded: Also for
Shrinkage cantilever beam, K = 1/4 ~ = 1/2
simple beam, K = 5/48 ~ = 1/8 hinged-fixed beam, K = 8/185 , ~ = 11/128
K 2 = deflection constant for the precast beam dead load
~ = deflection coefficient for warping due to shrinkage or
k
temperature change -- see K for values of~
= distance of neutral axis from compression flange -- see Eq. (39), also kr = 0,85 - 0.45(As' /As)•
kr = reduction factor to take into account the effect of compres -
=
sion steel, movement of neutral axis, and progressive cracking in reinforced flexural members; and effect of nontensioned steel in prestressed flexural members, see k for values of kr
2 2 2 1 + e /r , where r =lg/Ag
xx
L
LA
M
= span length in general and longer span for rectangular slabs. Also used as a subscript to denote live load
= subscript denoting loading age
= bending moment. bending moment, formly loaded:
When used as the numerical maximum for beams of uniform section and uni-
cantilever beam , simple beam ,
hinged-fixed beam (one end continuous), fixed-fixed beam (both ends continuous).
(-) M = q L2/z (+) M = q L2 /8 (-) M = q L2 /8 (-)M=qL2/12
= maximum bending moment under slab dead load for composite beams
M 2 = maximum bending moment under precast beam dead load
M 10 = bending moment between symmetrically place diaphrams
Ms, Di = bending moment due to slab or slab plus disphram, etc., dead load
Mer = cracking moment
Mmax = maximum moment under service loads
m = modular ratio of the precast beam concrete, EsfEcs' at the time of slab casting. Also used as subscript to indicate measured values
n = modular ratio, Es/Eci' at the time of loading, such as at release of prestress for prestressed concrete members. Also usually used as Es/Ee for reinforced members
P = applied transverse load for load-deflection studies
P = applied transverse load corresponding to the cracking er moment, Mer
= maximum value of applied repeated transverse load in a cycle
xxi
pult
PG
PG cp
PL
PL l cp
PL cp
PL el
PL u
p
p'
Q
= applied transverse load corresponding to the ultimate strength of the beam
= prestress gain in percent of initial tensioning stress or force
= pres tress gain due to creep under slab dead load at time t
= prestress gain due to differential shrinkage at time t
= elastic prestress gain at slab casting
= total pres tress loss in percent of initial tensioning s tress or force
= prestress loss due to creep prior to slab casting at time t
= prestress loss due to creep after slab casting at time t
= prestress loss due to creep at time t
= prestress loss due to elastic shortening
= prestress loss due to steel relaxation at time t
= prestress loss due to shrinkage of concrete at time t
= total prestress loss at any time t
= ultimate prestress loss
= steel percentage, As/bd for cracked members, and As/Ag for uncracked members. Also used in percent in shrinkage warping equations
= compressive steel percentage, A~/bd for cracked members, and A~/Ag for uncracked members. Also used in percent in shrinkage warping equations
= differential shrinkage force - D A1 E 1/3. The factor 3 provides for the gradual increase in the shrinkage force from day 1, and also approximates the creep and varying stiffness effects.
xx ii
q
r
s
= uniformly distributed load
= radius of gyration, r 2 = lg/ Ag
= subscript denoting time of slab casting. Also used to denote steel. Also used as subscript to indicate sustained
load
sh = subscript denoting shrinkage
t
t
= total depth or thickness of section. Also subscript to denote time-dependent
= time in general, time in hours in the steel relaxation equation, and time in days in other equations herein
tLA = age of concrete when loaded, in days
u = subscript denoting ultimate value
w = unit weight of concrete in pcf
x = subscript to indicate distance as measured from the end of the beam -- see Eq. (35)
Yes = distance from centroid of composite section to centroid of slab
Yt = distance from centroid of gross section to extreme fiber in tension
= ratio of creep coefficient at any time to ultimate creep
coefficient, c/ cu
= ratio of creep coefficient at the time of slab casting to C u
= creep correction factor for the precast beam concrete age
when loaded
i3s = creep correction factor for the precast beam concrete age when the slab is cast for composite beams
Y s = ratio of shrinkage at slab casting to shrinkage at ultimate (referred to 7-day initial reading)
xx iii
y' = ratio of shrinkage after slab casting to shrinkage at ulti-s
(referred to 7-day initial reading) mate
I::. = deflection or camber
I::.. = initial deflection, camber 1
( l::.i ) 1 = initial deflection under slab dead load
( I::. i) lD = initial deflection due to diaphram dead load
( l::.i J2 = initial deflection under precast beam dead load
( I::. i )D = initial dead load deflection
( l::.i )F = initial camber due to the initial pres tress force, F 0
0
I::. DS = differential shrinkage deflection
I::. L = live load deflection
I::. t = total camber, deflection, at any time
I::. u = ultimate camber, deflection
(€sh)t = shrinkage strain in inches/inch or cm/cm, etc., at time t
( € ) = ultimate shrinkage strain in inches/inch or cm/cm, etc. sh u
cp = curvature
cpsh = curvature due to shrinkage warping -- see Eq. (16)
= curvature due to shrinkage warping of precast beam up to slab casting -- see Eq. (20)
* 1
= load ratio for repeated load studies -- see Eq. (40)
xxiv
1
Chapter 1
INTRODUCTION
1. 1 Statement of the Problem
As a result of the increased use of structural lightweight con
crete for precast bridge girders along with normal weight concrete
deck slabs, a need exists for a better understanding of the factors,
primarily time-dependent, that affect prestress loss and camber (in
the case of prestressed girders) and deflection (in the case of rein
forced girders) in composite beams of these materials. Of particular
interest in this study is the behavior of sand-lightweight (100% sand
substitution for fines along with lightweight coarse aggregate) and all
lightweight prestressed structures in relation to normal weight pre
stressed structures, and the effect of the composite slab on the ulti
mate loss of prestress and camber. The effect of composite slabs
on the deflection of reinforced concrete members is also included in
this study.
In order to complete a comprehensive study of the initial plus
time-dependent deformational behavior of non-composite and com-
posite structures, the load-deflection response of reinforced and pre
stressed members under single cycle and repeated cycle shQrt-time load
tests (with constant and increasing load levels) into the cracking range are
2
also included in this study. Twenty-four hour sustained load tests into
the cracking range are also studied.
1. 2 Objectives and Scope
The principal objective of this investigation is to evaluate
experimentally the time-dependent behavior of sand-lightweight and
all-lightweight concrete beams (pres tressed and reinforced), includ
ing composite beams, in order to present practical design methods,
and to give an indication of their accuracy for predicting loss of
prestress and camber (in the case of prestressed beams) and deflec
tions (in the case of reinforced beams).
The study is divided into three parts: a materials study of
the concretes themselves, a laboratory study of the behavior of both
non-composite and composite beams that included prestressed (15
beams) and reinforced (3 beams) beams, and the field measurement
of camber of pres tressed girders (5 girders) used in the fabrication
of a compos.ite bridge in Iowa. The minimum test period for the
laboratory beams is 6 months, although data is recorded for 1 year
for 3 of the beams. The test period for the bridge girders is 560 ·
days.
The laboratory prestressed concrete beams are designed in
five groups (3 beams in each group) to investigate the loss of pre
stress, initial and time -dependent camber, load-deflection behavior
3
(under single and repeated load cycles) and the effect of different slab
casting schedules. One group of 3 reinforced beams is used to inves -
tigate the initial and time-dependent deflection, load-deflection behav
ior after sustained loading, and the effect of different slab casting
schedules.
Results computed by the methods described for predicting ma
terial behavior and structural response are shown to be in good agree
ment with the control specimen data, the laboratory beam data, the
bridge girder data, and other published experimental data. Continuous
time functions are provided for all needed parameters, so that the gen
eral equations readily lend themselves to computer solution. Approx
imate equations are also included.
1. 3 Review of Literature
Shrinkage of concrete is its contraction due to drying and
chemical change. Various empirical equations are presented in the
literature ill• @l_, ill for predicting shrinkage strains. AC! Com
mittee 435 (.!1 has given a quantitative resume of available informa
tion on creep and shrinkage as applied to deflections of reinforced
concrete beams.
Concrete undergoes time-dependent deformations under the
action of sustained loads that are attributed to creep of the concrete.
The contributions of Lorman (21, McHenry (!1, Neville L?2_,
Ross @)_, and, Troxell, et al ffi are noted. Lorman and Ross
4
suggested the use of hyperbolic expressions for predicting creep
(used in this report in modified form). McHenry's concept of
"superposition technique for creep" is used in this report; for example,
in the case of creep under slab dead load. Neville's study of the
physical nature of creep is noted.
A number of creep theories and mechanisms of creep have
been reviewed by Neville (1), Ali and Kessler (_!_Q). and Meyers, et al
(!..!). Meyers and Neville (_g_) and Pauw and Chai (_!l) have summa
rized the primary factors that influence creep. The influence of the
size and shape of the member on creep and shrinkage was also
reported by Hansen and Mattock (14).
The principal articles referred to in this report on the subject
of creep and shrinkage of all-lightweight and sand-lightweight con
crete are those of Jones, et al (15), ACI Committee 213 (~),
Pfeifer (!1)• Christiason (!~),Schumann (!..2)• and this project(~).
Although the behavior of non-composite and composite pre -
stressed beams of normal weight concrete has been studied in
References (20) through (34), etc., (most of these referred to non-
composite beams only), it appears that no such investigation has
been made of composite pres tressed members of lightweight concrete.
Lofroos and Ozell (~.U were apparently the first to report
experimental results of time-dependent camber of prestressed con
crete beams. The specimens were two pairs of post-tensioned
5
normal weight non-composite beams under different prestress levels.
Branson and Ozell (23) examined experimentally the initial
plus time-dependent camber of both composite and non-composite
post-tensioned beams of normal weight concrete. Methods for cal·
culating camber were developed using certain experimentally deter
mined coefficients. The predicted results were in fair agreement
with the measured values. It was also concluded that camber tends
to reach an ultimate value relatively early compared to creep and
shrinkage, because of the offsetting effects of loss of prestress and
camber growth due to creep.
Corley, Sozen and Siess (24) discussed at great length the
reduced modulus method, the rate of creep method, and the super
position method in a study of the time-dependent camber of pre
stressed concrete beams. The rate of creep method was deemed
preferable on account of its relative simplicity. It was concluded
that time-dependent camber could be objectionably high, if there was
high stress gradient in the beam.
Sinno (27) in his study of lightweight non-composite pres tressed
bridge girders, concluded that hyperbolic functions can be used to
predict loss of prestress and camber (used in modified form in this
report). He also observed that camber tends to reach an ultimate
value relatively early as compared to creep and shrinkage.
6
Yang (28) in a recent study of lightweight non-composite pre-
stressed beams, concluded that creep under constant stress and
variable stress was proportional to the applied stress within limits
of about 40% of the ultimate strength.,
Methods used in this study for predicting loss of prestress
and camber were based in part on the papers of ACI Committee
435 (29)and Branson (.£2.), (~Q).
With respect to short-time deflection of prestressed members
under static and repeated loading, the works of Abeles (35) - (38}, - -Burns (39), Hutton (40), and Warawaruk, Sozen, and Siess ('.!.!._) are
noted. Abeles 1 work primarily deals with partially pres tressed mem-
bers under static and fatigue loading. In general, it is concluded that
maximum tensile stress of the order of the modulus of rupture of the
concrete may be permitted under working loads without any detrimen-
tal effects on the serviceability and safety of the prestressed
members.
Burns (21_) has presented a detailed analytical method for
obtaining the moment-curvature relationship for partially prestressed
beams. The study was limited to pres tressed concrete beams with-
out non-tensioned steel.
Warawaruk, et al (41) in a comprehensive study of noncom-
pas ite prestressed beams presented methods for the prediction of
deflections of prestressed members at the various loading stages.
7
This method, however, is too elaborate as a design procedure.
The procedure developed by Branson (.?.Q_), (j), (lQ_), (42)
for predicting the deflection of reinforced beams under single-cycle
loading and adopted for the 1971 ACI Code (~), and applied to pre
stressed beams by Shaikh and Branson (49), is extended in this study
to the prediction of deflections of both reinforced and prestressed
beams under repeated load cycles into the cracking range.
8
Chapter 2
DESCRIPTION OF EXPERIMENTAL INVESTIGATION
2. 1 Materials and Test Specimens
The details of the laboratory beams and bridge girders are
shown in Figure 1 and Tables Al and A2. The laboratory beams were
designed as follows:
Group A -- 3 non-composite beams with different prestress moments made of sand-lightweight concrete.
Group B -- 3 beams, two of which are composite beams. The beams are made of sand-lightweight concrete. The slabs (of normal weight concrete) were cast at 4 weeks and 10 weeks after the casting of the beams. The same prestress moment is used for the three beams.
Group C - - Same as Group B but with a different pres tress moment.
Group D -- Same as Group A but made of all-lightweight concrete.
Group E
Group F
Same as Group B but with a higher stress level.
3 reinforced (non-prestressed) beams, two of which are composite beams. The beams are made of sand-lightweight concrete. The slabs (of normal weight concrete) were cast at 4 weeks and 10 weeks after the casting of the reinforced beams. The same steel percentage is used for the three beams.
0 20 40 60 80 100 120 140 160 180 200 Time in days
Fig. 6 Creep coefficient vs time curves for lab concrete (Gps D, E, F)
600~~~~~~~~~~--~~--r---~~c-~--~~--~~--~~ Measured .::"' ..... I
.:: 0
·;;; -400 !:: >: Ol ..cl
~-S200
Gp D ---Gp E
Iii Group D & Group E ct Group
~==;!ct=1-qr-+---r--t--1L-#. Gp F
ell" ..!< oi Computed .:: " ..... ..c: by Eq. (8) ,. u o~--......J----__J_----...L----..L.----L---......J----__J_----~:.._-..:~.:.....:.-
r]J .s 0 20 40 60 80 100 120 140 160 180 200 Time in days
Fig. 7 Shrinkage vs time curves for lab concretes (Gps D, E, F)
16
to 15% higher than the initial tangent values. However, the computed
initial camber of the laboratory beams and bridge girders was in
agreement with the measured results (Table 4 }. Eq. 6, developed
in Reference (18), is considered satisfactory for normal weight,
sand-lightweight, and all-lightweight concrete.
3. 2 Creep and Shrinkage
The principal variables that affect creep and shrinkage are
outlined and discussed in Appendix B. The design approach pre-
sented herein for predicting creep and shrinkage refers to "standard
conditions" and correction factors for other than standard conditions.
Based largely on the data and information from References
and this project, the following design procedure (developed in this
project and Reference (.!..§.), and used in Reference (42 )}, is recom-
mended for predicting a creep coefficient and unrestrained shrinkage
at any time, including ultimate values. The general values suggested
for Cu and (€shlu should be used only in the absence of specific
creep and shrinkage data for local aggregates and conditions. How-
ever, the "time-ratio" part (right-hand side except for C and u
(€sh>ul of Eqs. (7) - (9) have been found (_!1} to apply quite generally.
As shown in References (.!..§.}and (42}, these general values of Cu
and (€sh>u refer to average values only. See these references for
ranges of variation.
17
Standard creep equation -- 3" or less slump, 40% ambient relative humidity, minimum thickness of member 6" or less, loading age 7 days for moist cured and 1-3 days for steam cured concretes
c u lO+t0.60
(7)
For the laboratory beam lightweight concretes (moist cured) of this project, the following values apply:
GrouE Load. Age
A, B, c 7 days D 7 E 9 F 21
Rel. Hum.
40% 50 50 50
Cu
1. 75 1. 87 1. 80 1. 63
For the bridge girder sand-lightweight concrete project -- Cu= 2. 15 for H = 40%. H was 70%. H = 70%, Cu= 0. 80(2. 15) = 1. 72.
(steam cured) of this From Eq. (12) for
General value suggested for all weights of structural concrete (both moist and steam cured concrete, types I and III cement) -- Cu= 2. 35 for H = 40%. From Eq. (12) for H = 70%, Cu= O. 80(2. 35) = 1. 88.
Standard shrinkage equations - - 3" or less slump, 40% ambient relative humidity, minimum thickness of member 6 11 or less
Shrinkage at any time after age 7 days for moist cured concrete
(8)
For the laboratory beams lightweight concretes (moist cured) of this project, the following values apply:
GrouE Ini. Read. Age A, B, c 7 days
D 7 E 9 F 21
Rel. Hum. 40% 50 50 50
( e:sh )u
650 x 10- 6 in/in. 540 510 385
18
General value suggested for all weights of structural concrete (both types I and III cement) -- (e:shlu = 800 x 10-6 in/in for H = 40%. From Eq. (13) for H = 70%, (€shlu = O. 70(800 x 10-6) = 560 x lo-6 in/in.
Shrinkage at any time after age l -3 days for steam cured concrete
For the bridge girder sand-lightweight concrete of this project - -(e:shlu = 560 x io-6 in/in for H = 40%. H was 70%. From Eq. (13) for H = 70%, (e:shlu = O. 70 (560 x 10-6) = 392 x 10-6 in/in.
(9)
General value suggested for all weights of structural concrete (both types I and Ill cement) -- (e: 8 h) = 730 x 10-6 in/in for H = 40%. From Eq. (13) for H = 70%, (e:~h)u = O. 70(730 x lo-6) = 510 x 10-6
in/in.
In Eqs. (7), (8) and (9), tis time in days after loading for
creep and time after initial shrinkage is considered.
Values from the Standard Eqs. (7) - (9) of Ct/cu and
(e:sh)t/(e:shlu are:
1 mth 3 mths 6 mths .!.L!:. 5 yrs
Ctf Cu, Eq. (7) 0.44 o. 60 0.69 0.78 0.90
(e:shlt/(e:sh)u, Eq. (8) 0.46 0.72 0.84 o. 91 o. 98
The lower creep and shrinkage for the concrete of this pro-
ject, as compared to the average or general values, was probably
due to the high concrete strengths attained. The computed (in Eqs. 7
and 8) and measured creep and shrinkage for the moist cured con-
crete of this project are shown in Figures 3, 4, 6 and 7.
19
Correction factors
All correction factors are applied to ultimate values. However, since creep and shrinkage for any period in Eqs. (7), (8), and (9) are linear functions of the ultimate values, the correction factors in this procedure may be applied to short-term creep and shrinkage as well.
For slumps greater than 3", see Figure B3.
For loading ages later than 7 days for moist cured concrete and later than 1-3 days for steam cured concrete, use
Eqs. (10) and (11) for the creep correction factors (_!_§).
Creep (C.F. )LA= l.25t~~ 118 for moist cured concrete (10}
Creep (C.F.)LA =I. l3t~~ 095 for steam cured concrete (II)
where tLA is the loading age in days. For example,
When tLA=lO days, mo. 20 30 60 90
cu. (C. F. )LA=O. 95, 0.87 o. 83 0.77 0.74
st. cu. (C. F. )LA =0. 90. 0.85 0.82 0.76 0.74
For shrinkage considered from other than 7 days for moist cured concrete and other than 1-3 days for steam cured concrete, determine the differential in Eqs. (8) and (9) for
any period starting after this time. For shrinkage of moist cured concrete from 1 day (used to estimate differential shrinkage in composite beams, for example}, use Shrinkage C. F. = 1. 20.
For greater than 40% ambient relative humidity, use Eqs. (12) and (13) for the creep and shrinkage correction factors (~), (43), (44 ).
Creep (C. F. }H = 1. 27 - O. 0067 H, H = 40%
Shrinkage (C. F. )H = 1. 40 - O. 010 H, 40% = H = 80% = 3. 00 O. 030 H, 80% = H = 100%
(12)
( 13)
20
where His relative humidity in percent. For example,
When H = 40"/o, 50 60 70 80 90
100
Creep (C. F. )H = 1. 00, 0.94 0.87 o. 80 0.73 D.67 0.60
Shrinkage (C. F. )H = 1. DO. 0.90 0,80 o. 70 0.60 D.30 o.oo
For minimum thickness of members greater than 6 11, see
Figure B3 for the creep and shrinkage correction factors, as a func -tion of length of drying and loading periods. For most design purposes, this effect (as shown in Appendix B) can be neglected for creep of members up to about 10" to 12" minimum thickness, and for shrinkage of members up to about 8" to 9" minimum thickness.
This method of treating the effect of member size was based
on information from References (14), (18), (44), and this project. - - -For large-thickness members, refer to the method of Reference (14),
and others, for relating size and shape effects for creep and shrink-
age to the volume/surface ratio of the members, etc.
Other correction factors for creep and shrinkage, which are usually not excessive and tend to offset each other, are described in Appendix B. For design purposes, these may normally be neglected.
21
Chapter 4
LOSS OF PRESTRESS AND CAMBER
4, 1 Relaxation Tests
Relaxation measurements were made for three different dia-
meter 7-wire pres tressing strands. The results agreed well with the
equation suggested in Reference (45), as can be seen in Figure 8.
It should be noted, however, that the relaxation of steel stress
in a prestressed member takes place under decreasing steel strain
(due to creep, shrinkage, etc.), rather than at constant length as in a
relaxation test. The loss of prestress due to steel relaxation is also
affected by slab casting (level of stress in steel is raised) in the case
of composite beams. Due to these effects and the practice of over
tensioning to counteract the relaxation that takes place between the
time of tensioning and effective bonding of concrete to steel (this
practice was assimilated in the laboratory beam tests, where it is
noted in Figure 8 that about 2% relaxation takes place in 24 hours,
for example), it is felt that about 75% of the steel relaxation in a
constant-length relaxation test should be used in prestressed concrete
Initial plus time-dependent strain distribution diagrams from concrete strains measured on the sides of the beams
Typical experimental pres tress loss determined for end section at 180 days
fsi = 172 ksi, Es = 27 x 103 ksi, Observed cone. strain at cgs =
1001 x 10-6 in/in, Loss from meas. strains= (1001x10-6)(27x10 3 )(100)/172 = Inc. in meas. loss due to laterial dis tr. (det. as 2. 5% of 15. 7p Meas. loss due to steel relaxation (75% of value from Figure 8) = Total experimental loss of prestress
15. 7% o .. 4 5.5
21. 6%
Figure 9. Determination of experimental loss of prestress
23
It was concluded in Reference (46) that steel relaxation is
probably insignificant beyond 100, 000 hours (11, 4 yrs), and that this
ultimate value might be taken as twice the value at 1000 hours (1. 4
mths ). The relaxation equation recommended in this paper is the
same time-function (Log t) as that of Reference (45 ), except reduced
by 25% in magnitude and incorporating the idea of Reference (46) that
the ultimate value be taken as twice the value at 1000 hours. This
results in an ultimate steel relaxation for pres tressed concrete of
7. 5%, as shown in Term (4) of Eq. (14). Although Term (4) of
Eq. (14) was suggested on the basis of relaxation studies of 7-wire
prestressing strands used for pretensioned specimens, it is felt that
this is valid even for post-tensioned specimens (see comparison of
loss of prestress and camber of other published data in Sec. 4. 7).
for the bridge girder ultimate value in the example herein) overesti-
mates (on safe side) Term (3).
Term (4) is the prestress loss due to steel relaxation,
Assumes Max. value= 7. 5% (at or above 105 hrs = 11. 4 yrs). In
this term, tis time after initial stressing in hours. This expression
applies only when f 8 /fy is greater than or equal to O. 55, in which fy
is the O. 1%-offset yield strength.
25
The camber is given by Eq. (15). It is suggested that an
average of the end and midspan loss be used for straight tendons
(laboratory beams herein) and 1-pt. harping, and the midspan loss
for 2-pt. harping (bridge girders herein).
where:
Term (1) is the initial camber due to the initial prestress
force after elastic loss, F0
• See Appendix D for common cases of
prestress moment diagrams with formulas for computing camber,
Here F = F. (1 - n f /f .), where f is determined as in o i c s1 c
Term (1) of Eq. (14).
Term (2) is the initial dead load deflection of the beam.
(lli)D = K M L 2 /Eci lg. See Notation for K and M formulas.
Term (3) is the creep (time -dependent) camber of the beam
due to the prestress force. This expression includes the effects of
creep and loss of prestress; that is, the creep effect under variable
stress. llFt refers to the total loss at any time minus the elastic
loss. It is noted that the term, tiFt/F0
, refers to the steel stress
or force after elastic loss, and the prestress loss in percent, PL
(as used herein), refers to the initial tensioning stress or force.
26
f . 6 Ft The. two are related as:
Fo PL)~
el f , and can be
bFt closely approximated by F =
0
0
Term (4) is the dead load creep deflection of the beam.
Term (5) is the live load deflection of the beam.
The deflection at any time for a non-prestressed reinforced
beam is given by Eq. ( 16 ).
where:
bt
(2) (3) ~
2 = - (6i)D - kr Ct (l>i)D - Kw <Psh L
Term (1) is the in.itial dead load deflection of the beam.
2 (6· )D = K M L /E . I • See Notation for K and M formulas.
1 Cl g
Term (2) is the dead load creep deflection of the beam. kr
takes into account the movement of the neutral axis. See Notation
for values of kr•
Term (3) is the deflection due to shrinkage warping.
( 16)
(6sh )f~<Psh L 2 See Notation for values of Kw; cp sh =. 7 (E:sh )tp l/
3 /t
where pis the steel percentage and tis the thickness of the member.
Term (4) is the live load deflection of the beam.
Unshored and shored composite beams at any time, including
ultimate values
Subscripts 1 and 2 are used to refer to the slab (or effect of
27
the slab such as under slab dead load) and precast beam, respectively.
The loss of prestress, in percent of initial tensioning stress,
for unshared and shored composite beams is given by Eq. (17),
(8)
~]100 DS f .
Sl
where:
(n f )(Ct c 2
(3)
6Fs+6Ft 12 - c )(1 - )-
s2 2F0 Ic
( 1 7)
Term (1) is the prestress loss due to elastic shortening.
See Term (1) of Eq. (14) for the calculation off . c
Term (2) is the prestress loss due to concrete creep up to the
time of slab casting. C is the creep coefficient of the precast beam S2
concrete at the time of slab casting. See Term (2) of Eq. (14) for
6 Fs comments concerning the reduction factor, ( 1 -
2 F ).
0
Term (3) is the pres tress loss due to concrete creep for any
period following slab casting. C is the creep coefficient of the t2
precast beam concrete at any time after slab casting. The reduction
factor, (1 - t:.Fs + 6Ft), with the incremental creep coefficient, 2 F 0
28
(Ct2
- Cs2
), estimates the effect of creep under the variable pre-
stress force that occurs after slab casting. The reduction factor
term was modified from previous references. The expression,
12/Ic• modifies the initial value and accounts for the effect of the
composite section in res training additional creep curvature (strain)
after slab casting.
Term (4) is the prestress loss due to shrinkage. See Term
(3) of Eq. (14).
Term (5) is the prestress loss due to steel relaxation. In
this term tis time after initial stressing in hours. See Term (4)
of Eq. (14) for the maximum value and limitations.
Term (6) is the elastic prestress gain due to slab dead load,
and mis the modular ratio at the time of slab casting.
MS, Die
Ig , Ms
0. refers to slab or slab plus diaphram dead
' 1
load, and e, Ig refer to the precast beam section properties for
unshared construction and the composite beam section properties
for shored construction.
Term (7) is the prestress gain due to creep under slab dead
load. Ctr is the creep coefficient for the slab loading, where the
age of the precast beam concrete at the time of slab casting is
considered.
29
Term (8) is the prestress gain due to differential shrinkage.
QYcsec PGDS = mfcd' where fed = Ic , and fed is the concrete stress
at the steel c. g. s. See Notation for additional descriptions of terms.
Since this effect results in a prestress gain, not loss, and is normally
small (see Table 3), it may usually be neglected.
The camber of unshored and shored composite beams is given
by Eqs. (18) and (19), respectively.
Unshored construction:
(1) (2)
( 5)
[-
{IF s + ( 1
Fo
(4)
+ (1 -
(6)
I2 - Cs ({11.)2 - (C - C ) ({I.)
2 t 2 S 2 l 2 IC
( 8) (9) ( 10)
where:
(3)
Term (1) is the initial camber due to the initial prestress
force after elastic loss, F • See Appendix D for common cases of 0
( 18)
30
prestress moment diagrams with formulas for computing camber,
( t,1.)F . See Term (1) of Eq. (15) for determining F .
0 0
Term (2) is the initial dead load deflection of the precast
beam. See Notation for Kand M formulas.
Term (3) is the creep (time-dependent) camber of the beam,
due to the prestress force, up to the time of slab casting. See Term
(3) of Eq. (15) and Terms (2) and (3) of Eq. (16) for further explana-
tion.
Term (4) is the creep camber of the composite beam, due to
the prestress force, for any period following slab casting. Again,
see Term (3) of Eq. (15) and Terms (2) and (3) of Eq. (16) for further
explanation.
Term (5) is the creep deflection of the precast beam up to the
time of slab casting due to the precast beam dead load.
Term (6) is the creep deflection of the composite beam for any
period following slab casting due to the precast beam dead load.
Term (7) is the initial deflection of the precast beam under
slab dead load. ( t. .) 1 = K M 1 L 2 /E I . See Notation for Kand 1 cs g
M formulas. When diaphrams are used, add to ( t.i)l:
L2 a2 ( 8 - 6), where M 1D is the moment between dia-
phrams, and a is L/4, L/3, etc., for 2 symmetrical diaphrams at
31
the quarter points, third points, etc., respectively.
Term (8) is the creep deflection of the composite beam due to
slab dead load. Ctl is the creep coefficient for the slab loading,
where the age of the precast beam concrete at the time of slab cast-
ing is considered.
Term (9) is the deflection due to differential shrinkage. For
simple spans, t;DS = Qy L 2/8E I, where Q = D A 1 E 1/3. See cs cs c
Notation for additional descriptions of terms. The factor 3 provides
for the gradual increase in the shrinkage force from day 1, and also
approximates the creep and varying stiffness effects @2_). This factor
3 is also consistent with the data herein and elsewhere. See Table 4
for numerical values herein. In the case of continuous members,
differential shrinkage produces secondary moments ( similar to
effect of prestressing but opposite in sign--normally) that should be
included.
Term (10) is the live load deflection of the composite beam,
in which the gross -section flexural rigidity, E I , is normally used. c c
Shored construction:
t;t = Eq. (18), with Terms (7) and (8) modified as follows: ( 19)
Term (7) is the initial deflection of the composite beam under
slab dead load. ( t;i)l = K M 1 L2 /Ecs Ic. See Notation for Kand
M formulas.
32
Term (8) is the creep deflection of the composite beam under
slab dead load= Ct1
( lli)1
• The composite-section effect is already
included in Term (7).
The deflection of ordinary reinforced composite beams of
unshared and shored construction is given by Eqs. (20) and (21 ).
Unshared construction:
( 1 ) ,-----J'---,
(4)
(2)
(5)
k (Ct r 2
- K cp W SS
2 L 2 - K ( ) w 'Psh - 'Pss 2 2
(7) (8) ( 9)
- k r
(3)
12 - C )( ti. )D I
s2 1 c
Term ( 1) is the initial dead load deflection of the beam.
2 ( t.1.)0 = KM L /E .I • See Notation for Kand M formulas.
Cl g
Term (2) is the dead load creep deflection up to the time of
(20)
slab casting. k takes into account the movement of the neutral axis. r
See Notation for values of k r
Term (3) is the creep deflection of the composite beam for
any period following slab casting due to the precast beam dead load.
33
Term (4) is the deflection due to shrinkage warping up to the
time of slab casting. See Term (3) of Eq. (16) for further explanation.
Term (5) is the deflection due to shrinkage warping for any
period following slab casting due to the shrinkage of the precast beam.
See Term (3) of Eq. (16) for further explanation.
Term (6) is the initial deflection of the precast beam under
slab dead load. ([1.) 1 = KML2/E I. SeeNotationforKandM 1 cs g
formulas. When diaphragms are used, add to ([Ii) 1 :
L2 a2 (B - b ), where M 1D is the moment between dia-
phragms, and a is L/4, L/3, etc., for symmetrical diaphragms at
quarter points, third points, etc., respectively.
Term (7) is the creep deflection of the composite beam due to
slab dead load. Ct is the creep coefficient for slab loading, where l
the age of the precast beam concrete at the time of slab casting is
considered.
Term (8) is the deflection due to differential shrinkage. See
Term (9) of Eq. (18) for further explanation.
Term (9) is the live load deflection of the composite beam, in
which the gross-section flexural rigidity, E I , is normally used. c c
Shored construction:
At = Eq. (20), with Terms (6) and (7) modified as follows: (2 1)
34
Term (6) is the initial deflection of the composite beam under
slab dead load. ( A.)1
= KM L 2 /E I . See Notation for Kand M l cs c
formulas.
Term (7) is the creep deflection of the composite beam under
slab dead load = C (A .) 1• The composite-section effect is already t1 1
included in Term (6).
It is suggested that the 28-day modulii of elasticity for both
slab and precast beam concretes, and the gross I (neglecting the
steel), be used in computing the composite moment of inertia, I , c
in Eqs. (17), (18), (19), (20), and (21).
Special case of "ultimate loss of prestress, camber, and deflection
For computing ultimate values of loss of prestress and camber,
Eqs. (22) - (29) correspond term by term to Eqs. (14) - (21), respec-
tively.
PL u
=
Loss of prestress for non-composite beams, as per Eq. (14):
(2)
(n f )C ( 1 c u
(4) ,----A----,
+ o. 075 f . J Sl
AFu --)
2F 0
(3)
(22)
35
Camber of non-composite beams, as per Eq. ( 15):
( I) (2) (3)
~ ~
( 6F 6F
6 = (6 i)F - (6 i)D t - F u + (1 __ u)c) (6 i)F u 2 F u
0 0 0 0
(4) (5)
~ ,--A..._.,
- Cu (6 i)D - 6L (23)
Deflection of non-composite non-pres tressed reinforced beams,
as per Eq. (16):
(1) (2) (3) (4)
~ ~ ,-"-----,
6 = - (6 i)D k C (6 .)0
K L2 - 6L (24) u r u i w cpu
Loss of prestress for unshared and shored composite beams,
as per Eq. (17): ( 1 ) (2) ( 3)
PL u
= [ (n f c) + (n f }{et C )(I 6F 6F +6F 1
2 - 2 Fs )+ (nf )(1-et )C (1- s u) l c s u c s u 2 F
0 0 c
(4) (5)
+ (c h) E /(1 + npk ) + O. 075 f . - (m f ) S U S S Sl CS
(7) ( 8)
12 - (m f )( (3 C ) -
1 cs s u c
~
- PG ] .!Q£ DS f .
Sl
(25)
A u
36
Camber of unshared composite beams, as per Eq. (18):
(1) (2) (3)
= (A .)F 1 0
(-A; s + (1
0
(4)
AF +AF
AF s
- --) Ct 2 F s
0
(
. AF - AF + - u s
F + ( 1 - s u )( 1
2 F - et )C) (A. )F 0
(5) ~
-etC(A.)2 s u 1
(9) (10)
s u 1 0 0
(6) (8)
- (1 - et )C (A.)2 s u l
l c
(26)
Deflection of unshared composite non-prestressed reinforced
beams, as per Eq. (20):
( 1 ) (2)
r--"----.
A = - (A 1. )D - k a. C (A . ) -
u rsu 1D
(5)
2 -Ky cpL
w s 1 u
- A L
(6)
(3)
k (1 - et )C (A.JD r 1!I u 1
(7)
k ~ C (A.) I -r s u i
(4) ~
12 2 l
-K y cp L w s u
c (8)
,---'--..
12 - ADS l
c
(2 7)
37
Camber of shored composite beams, as per Eq. (19):
6u " Eq. (26), except that the composite moment of inertia is used
in Term (7) to compute (6i}1
, and the ratio I2/Ic' is eliminated in
Term (8). (28)
Deflection of shored composite non-prestressed reinforced
beams, as per Eq. (21):
6 = Eq. (27), except that the composite moment of inertia is used u
in Term (6) to compute { 6i)l' and the ratio I2 /Ic, is eliminated in
Term (7). (2 9)
It is noted that Eqs. (14) - (29) could be greatly shortened by
combining terms and substituting the approximate parameters given
below, but are presented in the form of separate terms in order to
show the separate effects or contributions to the behavior (such as
due to prestress force, dead load, creep, shrinkage, etc., that
occur both before and after slab casting.
Grossly approximate equations:
Non-composite beams (prestressed) --
6 =6.+6.C(l u 1 1 u
6. = ( t:,. )F 1 1 0
(30)
Composite beams (prestressed)
c (1 + 2u) - n f + (£ h) E + O. 075 fs 1·J cs s u s
(31)
= A 1• + A. C
1 u
12 (-I ),
c
38
Non-composite beams (non-prestressed)
2 = -(A i)D - Cu ( Ai)D - Kw (<Psh)u L ' where
<Pshu = Y 8
(c h) /t, and K . is defined in Notation. s u w
Composite beams (non-prestressed) --
12 2* A ~ Ai + Ai Cu (-
1 ) - K (<Psh)u L , where
u c w
K is defined in Notation.. w
4. 3 Required Calculations and Summary of General Parameters
(32)
{33)
(34)
Continuous time functions are provided for all needed material
parameters (and for different weight concretes, moist and steam
cured), so that the equations herein readily lend themselves to com-
puter solutions. Certain other read-in data (such as for the effect of
behavior before and after slab casting--a , ~ , m, and AF /F ) s s s 0
are also included. The parameters related to material properties are
summarized below, so that for composite beam hand calculations for
example; in addition to the section properties, prestress force, F , 0
and concrete stresses, fc• fcs' the only calculations needed for com-
puting pres tress loss and camber are the initial camber, deflections - -
'~ The ratio 12 /le is dropped out for the shrinkage term to account for the cumulative effects of shrinkage - i.e., before slab casting, after slab casting and due to differential shrinkage. For values of
y s' see Section 4. 3.
39
The following loss of prestress ratios at the time of slab
casting and ultimate are suggested for most calculations:
t::.F /F for 3 wks to 1 mth between prestressing and slab s 0 .
casting = 0, 11 for Nor. Wt., O. 13 for Sand-Lt. Wt., O. 15 for All-Lt. Wt.
t::.F /F for 2 to 3 mths between prestressing and slab s 0
casting = O. 15 for Nor. Wt., O. 18 for Sand-Lt. Wt., O. 21 for All-Lt. Wt.
f::.F /F = O. 22 for Nor. Wt., O. 25 for Sand-Lt. Wt., 0. 31 u 0
for All-Lt. Wt.
Note that these are defined as the total loss (at slab casting
and ultimate} minus the initial elastic loss divided by the prestress
force after elastic loss. The different values for the different weight
concretes are due primarily to different initial strains (because of
different E's} for normal stress levels.
I
The following average modular ratios are based on fc = 4000
to 4500 psi for both moist cured (M. C.} and steam cured (S. C.} con-
crete and type I cement; up to 3-mths f~ = 6360 to 7150 psi (using
Eq. 2) for moist cured and 3-mths fb = 6050 to 6800 psi (using Eq. 4)
for steam cured, and for both 250 Kand 270 K prestressing strands:
40
Modular Nor. Wt. Ratio (w = 145)
M.C. S.C. At release of prestress n == 7.3 7.3
For the time bet- = 3 weeks, m= 6. 1 6. 3 ween prestressing 1 month, 6.0 6.2 and slab casting: 2 months, 5.9 6. 1
3 months, 5.8 6.0
SandLt. Wt.
(w = 120) M.C. S.C. 9.8 9.8
8. 1 8.3 8.0 8.2 7.9 8.2 7. 7 8.0
Es = 27 x 106 psi for 250 K strands, Es = 28 x 10 6
All-Lt. Wt.
(w = 100) M. C. S. C. 12.9 12.9
10. 7 10. 9 10. 5 10.7 10.3 10.6 10.2 10.5
psi for
2 70 K strands, CL refers to the part of the total creep that takes place s
to. 60 before slab casting (CLs =
0 60 , as per Eq. 7), and 13 ( = the 10 + t • s
avg. Creep (C.F. )LA from Eqs. 10 and 11) is the creep correction
factor for the pre cast beam concrete age when the slab is cast (under
slab dead load). See Eqs. (7), (8), (9), and the correction factors
herein, for suggested values for C and ( e: h) • u s u
The following may be substituted for normal weight, sand-
lightweight, and all-lightweight concrete (moist and steam cured,
and types I and III cement):
For the time bet- = 3 weeks, CLS = o. 38, 13 s = 0.85 ween prestressing 1 month, o. 44, 0.83 and slab casting: 2 months, 0.54, 0.78
3 months, o. 60, 0.75
The following may be substituted for normal weight, sand-.
lightweight and all-lightweight concrete (moist cured) and Types I
and III cement for composite non-prestressed beams.
41
(For 'beam in position' at 7 days)':'
For the time bet-ween 1beam in pos -ition' and slab casting
4. 4 Sample Calculations
= 2 weeks, 3 weeks, 1 month, 2 months, 3 months,
y = o. 29, y s s 1 o. 38,
o. 46, o. 63, o. 72,
= 0. 71 0.62 0.54 0.37 o. 29
The following numerical substitutions for ultimate lass of
prestress at midspan, using Eqs. (17), (25), and ultimate midspan
camber, using Eqs. (18), (26), with the general parameters given
herein, are made for the sand-lightweight, steam cured composite
bridge girders (with slab moist cured) of this project:
O. 4L-pt. from end, e (midspan) = 14. 3 in, e (end) = 6. 2 in, f . :: Sl
190,000 psi, F. = 867 kips, A = 4.56 in2 , Ag= 520 in2 , p = 0.00883, 1 s
lg= 108, 500 in4 , MD (precast beam) = 410 ft-k, IC = 334, 100 in 4
(using slab width divided by a factor of E t /E 1 b = 3. 42/3. 41 = s em s a
1. 00), Ms, Di (slab plus diaphram moment at midspan) = 630 ft-k.
Modulii of elasticity (using Eqs. 2, 4, and 6 for concrete):
E = 28 x 106 psi, as suggested for 270 K grade strands herein. s
~:~ The differentials are to be used when the beam is 'in position 1 at an age other than 7 days. Eg: For a slab cast at age of beam = 35 days with the beam in position at age = 28 days, the values of Ys and Ys are (O. 46 for 35 days - 7 days = 1 month minus O. 38
1 -- --for 28 days - 7 days = 3 w.eeks) = O. 08, and ( 1. 00 - O. 46) = O. 54, respectively.
42
Slab Ec = 3,41 x 106
psi, for f~ = 3500 psi, w = 145 pcf (Table A4).
Precast beam -- (see description of m and n in general parameters
section herein for concrete properties),
E /n 6 2.86 x 106 psi, E = = 28 x 10 /9. 8 = ci s
E = E /m = 28 x 106 /8. 2 = 3, 42 x 106 psi.
cs s
Using Fi' At' and It, as per Term (1) of Eq. (14) or (17) or
(25) f = 2467 psi. As per Term (6) of Eq, (16) or (21), f = 1006 c cs
psi. These concrete stresses refer to the midspan section. As per
Term (1) of Eq. (15) or (18) or (26), for camber, F = F. (1 - n f / 0 l c
fsi) = 758 kips, using fc = 2467 psi.
From the general parameters section: n = E /E . = 9 8· S Cl • '
for 2 months period between prestressing and slab casting --
m = E /E = 8. 2, a, = 0, 54, \3 = O. 78, llFs/F = O. 18; llF /F = s cs s s 0 u 0
o. 25.
From Eqs. (7) and (9), for H = 70%, C = I. 88, (€ h) = u s u
510 x l0-6 in/in,
Initial camber and deflection, and differential shrinkage
deflection:
( lli)F = 4. 09 in, as per Term (1) of Eq. (15) or (18) or (26). 0
( lli)2 = I. 74 in, as per Term (2) of Eq, (15) or (18) or (26).
( lli)l = 2. 26 in, as per Term (7) of Eq. (18) or (26). This
deflection is due to the slab and diaphram dead load.
43
ADS = O. 49 in, as per Term (9) of Eq. (18) or (26).
Solutions for interior girders
Ultimate loss of prestress at midspan using Eq. (25):
Bet. Computed Experi- General Eqs. ( 14), Gen. Eqs. Ult. Eqs. Approx. Beam Pres. Loss Just mental (17) with exp. param. ( 14 ), ( 1 7) (22 ), (25) Eq. (31) No. & Slab Before Loss at at 180d for Lab. B with exp. with gen. lwith gen.
Cast Slab Cast 180 days and 560d for bdg gird. par am. par am. pa ram.
Mid Ratio End Mid End Ratio Mid Ratio End Mid End Mid End Mid
a All losses are expressed in percent of initial stress. The ratios in the table are: Computed/ Experimental. See Footnote b, Table 3, for a description of experimental parameters.
b
c
d
e
The laboratory beams and bridge girders were prestressed at age 7-9 days and 2-3 days, respectively.
See Figure 9 for an example of the experimental loss determination. The 180 day and 560 day times in the table refer to times after prestressing.
The laboratory beam concrete strengths (for Gps. A-C) at release were well beyond the range specified for the general parameters; so then and m values for these lab. beams were computed separately. However, for the lab. beams of Gps. D and E, the suggested n and m values are used. Where general parameters are used, a correction factor is applied for rel. hum. only.
No approximate equation was given for non-composite beams for loss of prestress.
Beam No.
Al
A2 A3 Bl B2 B3 Cl C2 C3 Dl D2 D3 El E2 E3
eFl eF2 eF3
aTABLE 2
MEASURED AND COMPUTED MIDSPAN CAMBER&: DEFLECTION FOR LABORATORY BEAMS &: BRIDGE GIRDERS
bTime Comp. camber by d
Computed Ult. Camber Bet. Camber just Gen.Eqs. (15), (16) Gen. Eqs. Ult. Eqs. Approx.
1S3 2. OS 2.22 1. 08 6Sd 3. 10 3. 13 1. 02 0,2S 0.21 0.84 o. 17 o. 14 o. 14 1S4 2. 10 2.22 1. 06 6Sd 3,0S 3. 13 1, 03 o. 20 o. 21 1. OS o. 17 o. 14 o. 14 lSS 1.90 2, 14 1. 13 60d 2.9S 3,04 1. 03 -0. 02 0, 07 - o. 01 o. 14 o. 14 1S6 l,8S 2.27 1. 23 60d 2.92 3. 16 1. 08 o. 30 O,S4 cl, 80 a.so o. S l O,S3
a All camber values are in inches, Ratios are: Computed/Measured. See Footnote b, Table 3, for a description of experimental parameters. Also, see Sample Calculations for a description of general parameters.
b
c
d
e
See Footnote b, Table 1. Beams Fl-F3 were in position at beam age= 21 days.
Camber has Figure 18 ), is 0, 22".
been reduced from about 3" before slab casting to less than 1/2" after 1 year (see This ratio is large for the near zero camber, even though the difference in camber
See Footnote d, Table 1,
The camber of beams Fl, F2, and F3 being non-pres tressed reinforced beams are negative in magnitude, i, e., the values in this table for the beams (F 1, F2, F3) refer to deflections,
Beam No,
Al A2 A3 Bl B2 B3 Cl C2 C3 Dl D2 D3 El E2 E3
a, bTABLE 3
COMPUTED ULTIMATE LOSS OF PRESTRESS AT MIDSPAN, BY TERMS, FOR THE LABORATORY BEAMS AND BRIDGE GIRDERS,
USING THE GENERAL EQUATIONS (14) & (17) WITH EXPERIMENTAL PARAMETERS
Creep Creep El. Gain Creep Gain Due Total Loss, Elast, Loss Loss Shrink Relax Due to Gain Due to Diff. Eqs, ( 14). Loss Before After Loss Loss Slab to Slab Shrink (17)
aThe table is arranged in order of terms in Eq. (17). All losses are expressed in percent of initial s tress.
bThe experimental parameters used in the calculations for this table are shown in Tables A4 and AS and elsewhere herein for the lightweight concretes of this project. The slab shrinkage is shown here only. The correction factors given herein for age of loading, humidity, and member thickness (8" for Br. Gir.) are used where appropriate with the experimental parameters. The resulting creep and shrinkage factors used are:
Slab Shrink. (from day ~l used in comp. diff.str. (x 10 in/in)
c = u (e h) = s u
( e h) = s u
Laboratory Beams Gp. A, B, C Gp D ~
40% 50% 50% 1.75 l.87 1.80 650 540 510
_Qp_X 50%
1. 63 385
* * 470
(only for Gps. 440
B & C) 440
Bridge Girder 152-156
70% 1.62
352
330
Also see the Sample Calculations for a comparison with the general parameter results.
Bm No,
Al A2 A3 Bl B2 B3 Cl C2 C3 Dl D2 D3 El E2 E3 Fl F2 F3
Initial Camber due to Prest.
a, bTABLE 4
COMPUTED ULTIMATE MIDSPAN CAMBER, BY TERMS, FOR THE LABORATORY BEAMS AND BRIDGE GIRDERS, USING THE GENERAL
EQS (15), (16), (18) & (20) WITH EXPERIMENTAL PARAMETERS
Initial ccreep carob. ccreep carob. DL Crp Bm DL El def Crp def Defl. up to s 1. cast after s 1. cast defl. up defl. due to due to due to or shk. warp or shk. warp to slab after slab slab Bm,DL up to sl. cast up to sl. cast cast sl, cast DL DL
Initial Initial ccreep camb. ccreep camb. DL Crp Bm DL El def Crp def De fl Total
Bm Camber Defl. up to s 1. cast after sl. cast defl. up defl. due to due to due to Camber No. due to due to or shk. warp or shk. warp to slab after slab slab diff. using Eqs
Prest. Bm.DL up to sl. cast up to sl. cast cast sl. cast DL DL shk. (15), (16) (18), (20)
Figure 29 Computed and experimental midspan camber of beam reported in Reference (l!_)
00 00
89
difference between the computed and measured values of midspan
camber of 30%. It should, however, be noted that in spite of the
wide difference of 30%, the actual difference for the worst data point
is less than O. 18". Realizing that this difference between the com
puted and measured values of camber is for a girder of about 100'
span, the difference of 30% has only an academic significance.
4. 8 Summary of Results Reported by Others and Conclusion
On the basis of Figures 19 - 29, and the specific conclusions
made in section 4. 7, the following general observations are made
concerning the design method suggested in this report and the experi
mental results of University of Florida (~), University of Illinois
(24), Texas A & M University (27), and University of Missouri(~_!):
1. The use of the average general parameters and the
general Eqs. ( 14) and ( 17) is a reasonable means of computing the
loss of prestress for both composite and non-composite beams.
Either an underestimation (Figures 19, 20, 22) or an overestimation
(Figures 24, 25, 27, 28) may occur, depending on the difference
between the experimental and general values of the creep and
shrinkage parameters. However, the maximum scatter between
the computed and the measured values of loss of prestress was
+ 20% (using average general parameters) for these studies.
90
2. The use of the average general parameters and the
general Eqs. (15) and (18) is a reasonable means of estimating mid
span camber for both composite and non-composite beams. Either
an underestimation (Figures 21, 23, 2 9) or an overestimation (Figure
26) may occur, depending on the difference between the experimental
and general values of the creep and shrinkage parameters. The
maximum scatter, however, between the computed and measured
values of midspan camber is_± 30% (using average general parameters).
This maximum value of scatter occurs only in 3 of the 18 beams
studied (Figures 23, 29) and even in these cases, the difference
between the computed and the measured value of camber is less than
O. 18 ". The scatter between the computed and measured values of
midspan camber for the remaining 15 beams is _± 25%.
3. The procedure suggested in this report for the predic-
tion of initial camber is adequate.
4. Camber computations are more sensitive to the choice
of creep and shrinkage parameters than loss computations for non
composite beams. The reverse is true for the composite beams
because of the offsetting effects that may result in "near zero" camber
or deflection values after slab casting. These offsetting effects are
primarily due to the elastic and creep deflections due to the slab
dead load, and increased stiffness of the section on the one hand as
91
opposed to the reduced prestress force and its creep deformation on
the other.
5. The choice of the value of the initial modulus of
elasticity can affect the loss of prestress and camber (see results
of tests at the University of Illinois). In fact, the value of E . Cl
affects camber more than the loss of prestress.
6. It is reasonable to expect that the use of general
parameters along with the approximate Eq. (31) (for ultimate loss
of prestress) and Eqs. {30) and (32) (for ultimate midspan camber)
will result in values slightly higher than those obtained by the use
of the ultimate Eqs. (22) to (27).
92
Chapter 5
LOAD-DEFLECTION STUDIES OF PRESTRESSED AND REINFORCED CONCRETE BEAMS
5. 1 General
Increasing interest is being shown in the design of prestressed
concrete members that crack under working loads. Since substantial
cracking occurs under working loads in ordinary reinforced concrete
members, cracking at service load levels in prestressed concrete
members should be acceptable provided appropriate safety and ser-
viceability requirements are met.
This chapter is devoted to the study of prestressed concrete
beam deflections under a single load cycle (a single cycle is defined
herein as a continuously applied increasing load to failure at a static
rate) and repeated load cycles, and reinforced concrete beam deflec-
tions under increasing loads and 24-hour sustained cracking loads.
Both rectangular and composite T-beams are included.
The details of the test beams are shown in Tables Al and A2.
The concrete properties of the laboratory beams at the time of the
load-deflection tests are shown in Table A6. The laboratory beams
were tested as follows:
93
Groups A, B, and D: Single -cycle load tests for pres tressed beams
Group C:
Group E:
Group F:
Repeated load tests with constant load cycle for pres tressed beams
Repeated load tests with increasing load cycle for prestressed beams
Increasing load and 24-hour sustained load tests for reinforced beams
Observed midspan deflections shown in Figures 32-34, 37-48
refer to the position of the beam just before the application of the
transverse load. If the deflections from the positions of the beams
before prestressing are desired, the initial camber under prestress
and dead load and the time-dependent camber must be subtracted
from the deflections in Figures 32-34, 37-48. A two-point loading
system {Figure 30) symmetrical about the centerline of the beam was
used in all of the tests.
5. 2 Single Cycle Load Tests of Pres tressed Members
Deflection of uncracked members
The elastic theory can be accurately applied to concrete beams
as long as the concrete is not cracked. Distinct changes occur in the
behavior of concrete members after first cracking. After cracking,
there is a change in the distribution of bond and shearing stresses
and the load-deflection response changes sharply.
94
The determination of cracking loads can be based on the elastic
theory, assuming that cracking starts when the tensile stress in the
concrete reaches its modulus of rupture. The accuracy of the elastic
P/2 P/2
5 1 -6 11 5 1 -6 11
15'-0"
Figure 30. Two point loading for 'load-deflection' studies of laboratory beams
theory and also the modulus of rupture obtained from the usual bending
tests as being representative of the tensile strength of concrete in
bending has been questioned (48). However, most available test data
indicates that the use of the elastic theory up to cracking (determined
with the modulus of rupture) is sufficiently accurate.
For a pres tressed concrete beam without non-tensioned steel,
the cracking moment is given by:
Ftlg
AgYt
where Ft = Fi - l\Ft; Fi is the initial prestressing force and
l\Ft is total loss in prestressing force obtained by
using Eq. (14) or (17).
(35)
95
Ag = gross area of section
I = gross moment of inertia of section g
Yt = distance of tension fiber from cgc
' fcb = modulus of rupture of concrete.
Shaikh and Branson (49) indicated that the cracking moment
of prestressed concrete beams is (for all practical purposes) not in-
fluenced by the addition of non-tensioned steel. It was concluded that
Eq. (35) may be used to compute the cracking moment of prestressed
concrete beams containing non-tensioned steel in addition to pre-
s tressing steel.
Deflection of cracked members
Under cracked conditions, the behavior of prestressed concrete
members and ordinary reinforced concrete members is similar.
Since ordinary reinforced concrete members are invariably cracked
under working loads, most methods for computing these deflections
do _take into account the effect of flexural cracking in some form.
For this investigation, the method of Brans on ( _! )(2.2_)(2.!_)(42)
was used to compute the deflections of the test beams. The choice
of this method (Eqs. (37) and (38)) is based on favorable comments
from designers and on its indicated accuracy in the A Cl Committee
435 report(_!) on deflections of reinforced concrete flexural mem-
bers. These have been proposed for the 1971 ACI Code (50}(2.!_).
96
For an elastic homogeneous member subject to flexure:
M CD= EI (36)
The curvature, cp, at any section can be readily obtained using Eq.
(36), with the appropriate bending moment, M, and flexural rigidity,
EI, at that section. For uncracked sections either the gross, or,
more precisely, the uncracked transformed moment of inertia may
be us ed. Under cracked conditions, however, because of the varying
amount and extent of cracking, the flexural rigidity, EI, is not a
constant.
Theoretically one could evaluate zones for which the cracking
moment is exceeded and thus calculate the corresponding transformed
section moments of inertia along the length of the beams, based on
appropriate cracked and uncracked sections. With the flexural
rigidity known along the length of the beam, curvatures could be com-
puted using Eq. (36) and deflections obtained by the usual procedures.
Due to the complexity involved in relating the height of cracks,
spacing of cracks, etc. to the flexural rigidity of the member, mostly
empirical or grossly approximate methods have appeared in the liter-
ature for computing flexural rigidity, EI, under cracked conditions.
Based on a sizable number of tests on rectangular beams
(simple and continuous) and T-beams, Branson (50) has presented
an empirical expression for the effective moment of inertia at a
97
given section, Ieff• The expression was given in a form that includes
the effect of extent of cracking as:
(3 7)
where: Mer = cracking moment as defined by Eq. (35)
M = bending moment at the section where Ieff is desired
I = moment of inertia of gross section g
Icr = moment of inertia of the fully cracked section using
Eq. (39). See Figure 31.
An expression for an average effective moment of inertia for
the entire length of the simply supported beam under uniformly dis -
tributed load was also given by Brans on (~) as:
+ (38)
where: Mmax = maximum moment in the span.
It is to be noted that Eqs. (37) and (38) apply only when Mor
Mmax is greater than or equal to Mer; otherwise Ieff =lg. For con-
tinuous beams, the average of positive and negative moment region
values in Eq. (38) is recommended (42)(50)~).
98
b b(kd)
3 ·I , ..
I = +nA (d-kd)2
kdl __
er 3 s
(np)2 k = + 2np - pn d A
n = E /E s where: --- s c
p =A /bd s
Figure 31 Moment of inertia of cracked section (I ) er
-·-.,,
_, .,,
{39)
The concurrence of AASHO, ACI, and PCI codes on the methods
of determination of ultimate strength of prestressed concrete beams
establishes the reliability of the equations indicated in the codes.
Therefore, in this investigation only a comparison of observed and
computed (using equations from the ACI code) values of ultimate load
was obtained.
Single cycle load tests were conducted on all the beams of Grps.
A, B, and D. Midspan deflection of the test beams were obtained up to
loads ranging from 76 to 88 percent of the ultimate loads. Eq. (38)
''The same equations are also valid for composite beams (with trans -formed compression flange width to account for the different concretes) if the neutral axis falls within the flange. This was the case for the laboratory composite test beams studies herein.
99
was used to determine the effective moment of inertia in the computa-
tion of deflections. Eq. (35) was used for computing Mer• and Eq.
(39) was used for the determination of Icr· The modulus of rupture,
I fcb' was obtained by bending tests on plain concrete specimens for
the test beams. It is observed that Eq. (38) was originally established
for use in the case of simply supported beams under uniformly dis-
tributed loads. Its use, however, is considered adequate for the two-
point test loading.
The comparison of observed and computed midspan deflection
curves are shown in Figures 32 to 34. Table 5 shows the computed
and measured values of ultimate loads as well as the maximum dis -
crepancies in the observed and computed deflection curves.
Based on Figures 32 to 34 and Table 5, the following cibserva-
tions are made:
1. There are three distinct stages of behavior in the load-deflec-
tion history of a prestressed concrete beam. In the first stage, the
curve is virtually linear. This stage represents the behavior of the
beam before cracking of the concrete. The extent of this stage
depends on the geometrical and material properties of the section
and the type of loading. In the second stage, the load-deflection
curve is characterized by a constantly changing rate of deflection
with applied load and represents the behavior of the beam after the
e Worst discrepancy in -12% -12% -21% -24% -14% -13% -5% +8% deflection curves
D3
6. 83
6.84
3. 17
2. 15
5.25
77%
5.00
-10%
aThe computation of ultimate loads is based on accepted procedures indicated in ACI 318-63 Code. The corresponding equations are not reproduced here. The test period varied between 45-60 min for each beam.
TABLE 5 (Cont'd)
bFor the test beams, the working load was assumed to represent the condition that cracking would occur as soon as this load was e:i<ceeded. These values of P w were the computed cracking loads.
c Represents the maximum load for which deflections were recorded.
d Represents the load at which the discrepancy between the observed and computed deflection is the greatest.
ePlus or minus indicate that computed deflection is greater than or smaller than the observed deflections.
105
concrete is cracked and while the reinforcement stress is still in the
'elastic' range of the stress-strain curve for the reinforcement. The
third stage is marked by a very slow change in the slope of the load
deflection curve. In this stage, the reinforcement stress is in the
'inelastic' range of the stress-strain curve for the reinforcement
and the load-deflection curve is nearly flat.
In addition, the presence of non-tensioned steel affects the
deformational behavior of a prestressed concrete beam after the
initial cracking (49). It was concluded by Shaikh and Branson (49),
that the~ deflection in a beam with non-tensioned steel as compared
to the deflection of an identical beam without non-tensioned steel may
be greater, comparable, or considerably smaller depending on
whether the applied transverse load is approxim.ately equal to, some
what greater than or considerably greater than the cracking load.
Failure of the beam is usually the res ult of failure of the
compressed concrete. However, a beam with a very small percentage
of reinforcement may fail by fracture of the reinforcement. The third
stage, however, is not exhibited by beams having a high value of
steel percentage. The first two stages described above can be seen
clearly for the laboratory beams in Figures 32 to 34 (the steel percent
age varied from 0. 93% to O. 38% for rectangular beams and was of the
order of O. 1 % for the composite beams).
106
2. The level of prestress affects the shape of the load-deflection
curves. An increase in the level of prestress tends to increase the
load required to produce the flexural cracking and thus extends the
first stage. For example, Beam Al (whose prestress level is
greater than that of either Beam A2 or Beam A3) has a cracking load
of 3. 57k as compared to 3. 04k for Beam A2 and 2. 67k for Beam A3
(see Figure 32 and Table 5).
3. It is observed that for most of the beams (8 out of 9) studied
under single load cycle (see Table 5 ), the computed values of deflection
are smaller than the observed values of deflection. It is also observed
that the discrepancy between the computed and measured deflection
curves increases as the applied transverse load approaches the ulti
mate load capacity of the beam. Realizing that the tendency of con
crete to creep under load exists even for very rapid rates of loading
(52), it may reasonably be assumed that the discrepancy between the
computed and observed deflection curves is due to the creep of con
crete, Each load cycle required about 45-60 minutes to complete,
This creep effect has not been accounted for in the development of
Eq. (38). No attempt, however, is made to modify Eq, (38) for
creep effects, because the use of Eq, (38) gives reasonable estimates
of deflection (from a design point of view) up to 1, 5 to 2. 0 times the
working load.
107
4. The use of Eq. (38) resulted in computed deflections being
slightly greater than the observed deflections in most of the beams
(8 out of 10) in Reference (49), while in the current study the use of
the same equation results in the computed deflections being slightly
smaller than the observed deflections. This effect appears to be due
to the presence of non-tensioned steel in the beams reported in Refer
ence (49) which tends to reduce the creep effect and to further distri
bute the cracks along the beam.
5. The composite Beams BZ and B3 exhibit greater resistance
to applied loads than non-composite Beam Bl due to the inherent in
creased stiffness of the former (see Figure 33 and Table 5).
6. There does not seem to be any significant difference in the
load-deflection response of composite beams for which slabs have
been cast at different times. Both Beams BZ and B3 have almost
identical load-deflection curves (see Figure 33). However, there
could be a significant difference in the ~deflections (when referred
to the position before prestressing) due to the difference in the time
dependent contribution to camber (see discussion in Chapter 3).
5. 3 Repeated Load Tests of Pres tressed Members
Under single cycle loading, the load-deflection response of
prestressed concrete members can be reasonably predicted in both
the 1uncracked' and 'cracked' stage. This has been discussed in
108
Section 5. 2. However, under repeated loading, the 'load-deflection'
response is different.
To understand clearly the effect of repeated loads on pre
stressed concrete beams, it is necessary to know the effect of repeated
loads on the two components of pres tressed concrete, i.e., plain con
crete and pres tressing steel. Shah and Winter (22_) studied the behavi
or of plain concrete prisms with flared ends under uniaxial compres -
sion, cycled at stress levels below the ultimate strength of the prism.
They found that concrete possessed a shakedown limit at around 88 to
95 percent of the ultimate load. Below this level, concrete is rela
tively insensitive to several cycles of loading. Neither the strength
nor the strain capacity is affected below the shakedown limit. Pre
stressing steel like reinforcing steel, behaves (for all practical pur
poses) like elasto-plastic material. Repeated loading at load levels
below the yield strength of the material results in full recovery,
while above the yield strength of the material results in an 'inelastic'
set.
In this study of prestressed concrete beams, it is assumed
that under repeated loading, the stress in concrete is below its
'shakedown limit' and the stress in steel is below the 'yield strength'
of the steel. This implies that (1) if_ the concrete stress at the
repeated load level is· below the shakedown limit and (2) if the steel
109
The determination of deflections in the uncracked region (OA
in Figure 35} and cracked region (ABC in Figure 35) has been dis ...
cussed in Section 5. 2, The use of Eq. (38) implies the determination
of the point Bon the assumption that the slope of OB is proportional
to the effective moment of inertia, Ieff' The reliability of this equa
tion has been accepted (~Q_)(~ __ !). Unloading from the point B along BD
(a line parallel to OA) indicates that there is only elastic recovery.
This is true if the beam is severely cracked. If, however, the beam
is not severely cracked a certain number of cracks will close on
unloading (especially in regions of moments close to the cracking
moment). This will result in a small amount of 'inelastic' recovery.
This is indicated by FD in Figure 35. It follows, therefore, that the
total recovery (FE in Figure 35} is a function of the cycling load ......
the closer the cycling load is to the cracking load, the greater will
be the total recovery. This is also a logical extension of the fact
that when the beam is completely uncracked, the total recovery
(indicated by EF in Figure 35) is equal to the total deflection.
On the basis of the above discussion, the following relation-
ship is suggested for computing the average effective moment of
inertia under repeated loads:
= (40}
where:
110
stress (in the same concrete member at the same load) is below the
yield strength of the steel, then the reloading curve after attaining
the magnitude of the repeated load will follow the single cycle load-
deflection curve as if nothing else had happened (see Figure 35). In
Figure 35, this is indicated by the fact that if OAC is the single cycle
load-deflection curve, and if cycling is done at a load corresponding
to OB', the reloading curve (FB) will reach the point B and will follow
BC as if nothing else had happened.
B'
0 F D E
Deflection, 6
c
Slope of OB is proportional to 1eff
Slope of BF is proportional to I rep
Slope of BD is proportional to
lg
Figure 35 Details of deflections under repeated loadings
111
I is used to compute the recovery during the unloading rep
part of the cycle. (Note that the slope of FB is proportional
to I in Figure 35,} rep
= (40-a)
Ieff = effective moment of inertia as defined by Eq. (38)
lg = gross moment of inertia
Pult = estimated ultimate load based on current ACI procedures in the code
p er = load at initial cracking corresponding to Mer
(using Eq. (35)).
Prep= cycling load or maximum load in a given cycle.
Eq. (40) is valid only if the loading cycle produces cracking,
i e P ) P The value of *l requires some explanation. • · ' rep er·
From Figure 35, it is clear that the slope of the line BF is greater
than the slope of the line OB, but smaller than the slope of the line
BD. Also, the slopes of.lines OB and BD are proportional to Ieff
and lg respectively. For a severely cracked beam (Prep; Pult),
the total recovery consists of only the elastic part of the deflection
corresponding to the magnitude of the repeated load, Prep' For an
uncracked beam (Prep; Per)' the total recovery is equal to the
deflection corresponding to the magnitude of the repeated load, Prep•
The value of * 1 interpolates linearly between the two limits described
above. For example:
112
(a) when Prep = Per' '1Ji = 1, and !rep = Ieff = lg. This is a
(b)
condition of total recovery.
when P ' P , ~ 1 varies between 1 and zero, and !rep is rep/ er
between Ieff and lg• This a condition of some inelastic
recovery due to the cracks being closed.
(c) when Prep = P ult' W 1 = 0, and !rep = lg. This is a condition
of no inelastic recovery due to the cracks being closed. This
may also be considered as a condition of maximum residual
deflection.
Thus, the use of Eq. (40) enables one to predict the effective
moment of inertia under repeated cycles for any given range of
loading.
Also, the use of Eq. (40) in determining the effective moment
of inertia under repeated loading allows the slope of BF (see Figure
35) to become proportional to !rep·
In the development of the relationship in Eq. (40), the follow-
ing are implicitly assumed:
1. Absence of hysterisis loop in the unloading-reloading
sequence.
2. Absence of time-dependent effects due to creep during
the test.
The first assumption is justified on the basis that the stresses
113
due to repeated loading in concrete and steel are well below the shake-
down limit and the yield strength of the concrete and steel respectively.
This has also been observed in the study of reinforced concrete beams
under repeated loading in similar loading regimes (54). The second
assumption is probably justified on the basis of the small time involved
in the tests (see Table 6).
In this work, repeated loads mean a small number of cycles
at loads ranging from 1. 05 to 1. 43 times the working load (this cor-
responds to 55 to 72% of the ultimate load}. The working load is
defined herein as the load at which flexural cracking is initiated. The
following sample calculations indicate the use of Eq. (40) in the deter-
mination of deflections of prestressed concrete members under re-
peated loading.
Sample calculations for the deflection of a prestressed concrete beam under three cycles of loading
To illustrate the procedure outlines above, the midspan deflec-
tion of beam E 1 is computed under three cycles of repeated transverse
loads of the following magnitude:
(1) Prep = 5. 0 kips in the first cycle
(2) Prep = 5.5 kips in the second cycle
(3} Prep = 6. 0 kips in the third cycle. Note that P rep
corresponds to the maximum load in a specific load cycle.
114
The equations needed for the computations are Eqs. (35), (38),
(40), and (41). The pertinent geometrical and material properties
for the example beam are shown in Tables Al-AZ and A6.
For a simply supported beam under a two-point symmetrical
loading system (see Figure 30), the midspan deflection, A, is given
as:
= pa (8a2 + 12ab + 3b2 ) 48 EI
(41)
where: b = distance between the loads
a = distance of each load from the near support
P = total load on the beam
E = elasticity modulus of concrete
I = moment of inertia.
Referring to Figure 30, a = 5. 5 ft; and b = 4. 0 ft. For purposes of
illustration, the computed deflections will be referred to Figure 36.
Deflection, A
Figure 36 Sample Calculations
OABE, EBCF, and FCDG repre-
sent the first, second, and third
cycle respectively. The values of
Prep correspond to OB', OC 1, and
OD' during the first, second, and
third cycle respectively. OA'
represents the 'cracking load'
(also referred to as the working
115
load) and is defined as the load at which flexural cracking is initiated.
Parameters and terms for beam El
Span= 15'; e (midspan}= e (end}= l. 75"; F = 38, 7 kips (deter-
mined as Fi - ti Ft' where Ft is obtained using Eq, (17) in Chapter 4};
Figure 38 Observed and computed midspan deflection versus load curve of beam C2 under 3 cycles of repeated loading (one non-composite prestressed beam)
Figure 39 Observed and computed midspan deflection versus load curve of beam C3 under 3 cycles of repeated loading (one composite prestressed beam)
8. 0 Maximum value of repeated load, Pr= 5. ok, 5. 5k and 6. ok
6. 01---~
"Cl cd 0 10. 0 ~
8.0
6.0
4.0
2.0
0
• 2
----
6
•
0 0.3
• 4 0
. 6
. 2
Computed
in cycles 1, 2, and 3, respectively
• 8 • 4
1. 0 . 6 0
. 8 1. 0 2 4
(loading and unloading)
1. 2 1. 4 6 R 1 0
Observed (loading and unloading) Loading _/.'::,, --
f- )- ~-Unloading ~--
- --r- -
;;/ IF
Crack. loac Ulti. Bm Type
Com;' rorn~ Meas
/ El Re ct 4. 15k 4. 2rf 8. 55k
v " • 6 . 9 1. 2 1.5 1.8 2.1 2.4 2.7 3.0
Midspan deflection in inches
1 ? 1 4 1. 6 1. 8
---
load ''For details of compu-
Meas tation,
8. 3 lk refer to text
3.3 3.6 3.9 4.2
Figure 40 Observed and computed midspan deflection versus load curve of beam El under 3 cycles of repeated loading (one non-composite prestressed beam)
0
20.0
16.0
12.0
8.0
4.0
0
Maximum value of repeated load, Pr= 10. Ok, IO. 5k & 11. Ok in cycles 1, 2 and 3, respectively for Beams E2 and E3
Figure 41 Observed and computed midspan deflection versus load curves for beams E2 and E3 under 3 cycles of repeated loading (two composite prestressed beams)
':":'Only one curve is shown for the computed and measured values (for both beams E2 and E3) because only very small differences in deflection existed between the two beams.
-N N
123
3. 0 ' ., " " u
p.. p.. II II
'O 'O
"' "' 0 0 - -@ @ . 4-< 4-< ., ., 'O 'O
'O 'O ., ., > > " " ., ., "'
., 1. 0
& Cl (Rect) C!!IE 3 ( T Bm) ( 1 19 - 1 4 3 P )
O C2 (T Bm) • 1 • er
~-·> e C3 (T Bm)
o E 1 (Rect) ...... • E2 (T Bm) ..-.- _>-··_,.
-··- C!!I --- 1, 04-1, 15 P~~) .......,; Ill.,.--- _....d: 0 8 pc r -.. .. ··1 1-• - -
I
2. 0
..0 ..0 0 0 1.06 per - -"' "' ...., ...., 0 0
[-< [-< 0
0 1 2 3 4 No. of cycles
Figure 42 Effect of repeated loading (in the cracked range) on total deflections of laboratory beams of Groups C and E
b
TABLE 6
aDETAILS OF REPEATED LOAD CYCLES AND DISCREPANCY IN THE OBSERVED AND COMPUTED VALUES OF MIDSPAN DEFLECTION
FOR BEAMS OF GRPS C & E
c d e Comp Meas Work Load Cycling Ld, Comp. res. def. Meas. res. def.
Detail ult, ult. load, factoi p pr (total)@ end of (total)@ end of ld, ld, PW Pu/
max for cvcles cvcles cycles
Pu Pum Pw 1 2 3 1 2 3 1
u Cl 7.98 fl, zo 3.67 2, 18 7.00 4.5 4.5 4.5 . uo83 • Ob83 • 0685 • 0600
a All loads are expressed in kips and all deflections are expressed in inches. b
See Footnote 1, Table 5, c
See Footnote 2, Table 5. d
See Footnote 3, Table 5,
2 , uo I 0
• 0040
• 0040
. 2310
• 0090
• 0080
3 • uo!O
. 0040
• 0040
.480
• 025
• 026
f Worst disc rep. in defl. curves
+ 14o/o
-18%
-34%
+10%
-33%
-33%
eThe magnitudes of residual deflections being very small, any meaningful interpretation on the basis of a percentage of deflection at working load, (say Pw) is difficult. See Sample Calculations also.
f The discrepancy in the deflection curves refers to the load-deflection curves after the cycling loads have been completed, The high values of discrepancy in this column corresponds to about 80-82% of the ultimate load. Also, see Footnote 5, Table 5.
125
the repeated load in a specific cycle, P ) and the number of cycles rep
at various load levels. Table 6 shows the computed and measured
values of ultimate load as well as the computed and measured mag-
nitudes of residual deflection.
Based on Figures 37 to 42, Table 6, and the sample calculation,
the following observations are made:
1. The residual deflection at the completion of a cycle is a func-
tion of the load at which the cycling is done. At cycling loads close to
the ultimate load, the residual deflection is larger than at cycling
loads close to the cracking load (see Table 6 and Figures 37 to 41).
2. Repeated cycles (up to three cycles) of loading at a given load
level does not increase the magnitude of the residual deflection.
Similar observations have been made on reinforced concrete beams
(54) under load levels below the yield strength of the reinforcement.
3. The magnitude of the total recovery decreases with increasing
load (see sample calculations). This was the basic premise on which
Eq. (40) was developed, and is confirmed by observations (see Table
6).
4. The residual deflection at the end of a cycle is also a function
of the geometric properties of the section. This, though obvious, is
clearly seen in Figures 37 and 38, where the .composite beams have
less residual deflection than non-composite beams even at the same
level of loading.
126
5. There does not seem to be any significant difference in the load
deflection response of composite beams under repeated loading for
which slabs have been cast at different times. Both beams E2 and E3
have similar magnitudes of residual deflections and ultimate loads
(see Figure 41 and Table 6).
6. It may safely be concluded, that the relationship suggested by
Eq. (40) gives reasonable agreement between observed and computed
values of deflections provided the stress under repeated loading in
concrete and steel are below the shakedown limit of the concrete and
the yield strength of the steel respectively. In the case of the labora
tory beams, the range of the repeated load varied between 55 to 72%
of the ultimate load. This corresponds to 1. 05 to 1. 43 times the
working load. It is reasonable to expect that as the repeated load
approaches the working load, the total recovery approaches the total
deflection. This has been discussed in detail elsewhere. Also, com
parison with data in the literature confirms the use of Eq. (40) as a
reasonable means of estimating the effective moment of inertia under
repeated loads for reinforced concrete beams under similar loading
regimes (see Section 5. 5 ).
7. It is reasonable to expect that at repeated loads close to the
ultimate load (yield of steel reinforcement in the case of under-rein
forced beams), there will be greater residual deflection (than when
127
steel has not yielded) as well as a hysterisis loop during the loading-
unloading sequence. A detailed study of reinforced concrete beams
in this loading regime has been reported by Ruiz (55).
5. 4 Increasing Load Plus 24-Hour Sustained Load Tests
Although much work has been reported on the effect of sus -
tained load on reinforced concrete beams (i_), (2.Q_) most of these works
referred to beams at early loading ages. In this study, the beams
were loaded (at beam age 6 months) into the 'cracked' or 'inelastic'
range and left in that position for 24 hours.
Increasing load plus 24-hour sustained load tests in the
cracked range were conducted on beams of Group F (one non-composite
and two composite members). Midspan deflections on all the test
beams were obtained up to loads ranging from 79 to 92 percent of the
ultimate loads. The sustained loads ranged from 33 to 92 percent of
the ultimate loads. Eq. (38) was used to determine the effective mo-
ment of inertia, Eq. (35) was used to determine the value of Mer
I
(with Ft= 0). The modulus of rupture, fcb' was obtained by bending
tests on plain concrete specimens of the test beams.
The creep coefficients for computational purposes were based
on information and test results presented in Chapter 3. The following
sample calculations indicate the use of Eqs. (7), (38) and the appropriate
creep coefficients in the computation of deflections of reinforced
128
concrete members under increasing load plus 24-hour sustained
loading.
Sample calculations for the deflection of a reinforced concrete beam under 24-hour sustained loads
Beam Fl is selected for illustrating the calculation of deflection
under 24-hour sustained load in the 'inelastic' range of the load-deflec-
tion curve.
Parameters and terms for Beam Fl:
Span = 15 ft; e (midspan) = e (end) = 2 in; A = O. 6 in2; Ag = 48. 0
fcb = 430 psi (see Table A6); fc = 4540 psi (see Table A6); Pult (based
on ultimate equations given in ACI 318-63 Code) = 3. 56 k; age of beam
at load deflection test= 201 days
Deflection under sustained load, Psust = l. 2 kips
~ax= MDL+ Mtransverse load
Ieff (using Eq. {38))
lli (using Eq. (41))
= 13. 1 + 1. 2 x 5. 5 x 12 I 2
= 53. 3 inkips
124.lin4 =
= O. 356 in
as compared to the observed value of O. 350 in {see Figure 43),
Experimental C (from 7 days at u
40% RH)
Correction factor for 50% RH (using Eq. (12))
= 1. 95
= 0.94
Correction factor for age of loading (using Eq. (10))
Actual Cu
''Experimental value of Ct/Cu at 1 day (based on loading at 9 days)
Actual Ct for 24-hour loading
':":'Deflection due to sustained loading (using Term (2) of Eq. (16))
129
= o. 684
= 1.95 x o. 95 x 0.684 = 1,26
= 1/8
= 1.26 x 1/8 = 0.158
= • 158 x • 356 x • 85 = • 048 in
as compared to the observed value of 0. 053 in (see Table 7)
''The experimental value of Ct/ Cu and not the computed value (based on Eq. (7)) is used in the calculations, because the validity of the latter for extremely short periods is questionable, although the equality of this ratio (Ct/Cul for various loading ages is implicitly assumed in the equations for creep (see Chapter 3).
':":'The effect of shrinkage is very small (due to the very late age of loading as well as the short time period of the test) and is considered negligible.
The comparison of observed and computed midspan deflection
curves are shown in Figure 43. Table 7 shows the computed and
measured values of the ultimate load, as well as the computed and
measured values of the deflections due to the 24-hour sustained load.
Based on Figure 43 and Table 7, the following observations
are made:
1. The magnitude of the deflection due to sustained loading (24
hours) is a function of the level of the sustained load. Beam F2 has
Figure 43 Observed and computed values of midspan deflection for beams of Group F under 24-hr sustained loading (one non-composite and two composite reinforced beams)
.... "' 0
Detail
Fl r..,
"' F2
'" 0 F3
TABLE 7
aDETAILS OF INCREASING LOAD PLUS 24-HR SUSTAINED LOAD TESTS WITH REGARD TO WORKING LOADS, ULTIMATE LOADS AND
DEFLECTIONS UNDER THESE LOADS
b e Working Ult. Load Load Sustained Def. due to Sustained Worst load, c Factor Ld. factor load Discrepancy
PW Comp Meas Pu/PW p /P d Meas in def. curves s w Comp
0.42 3.56 3.62 8.5 2.86 • 048 • 053 +10%
1. 11 5.39 5.42 4.9 4.50 . 062 . 065 +5%
1. 11 5.41 5.61 4.9 2.88 • 032 . 032 +5%
a All loads are expressed in kips and all deflections are expressed in inches. The period of the test varied between 15-25 min for each beam prior to the application of the sustained load and between 10-20 min after the end of the sustained load.
b See Footnote 2, Table 5.
c See Footnote 1, Table 5.
dThe creep coefficient was the experimental value of Ct for the 24-hour sustained loading. See Sample Calculations.
e See Footnote 5, Table 5.
132
a higher deflection under sustained load than Beam F3 due to the higher
level of loading due to the higher level of loading in the former (see
Table 7 and Figure 43). The deflections due to creep is approximately
proportional to the applied load.
2. For extremely short periods of sustained loading (24 hours),
the use of Eq, (7) for the determination of creep coefficients in the
computation of deflections due to sustained loads is, perhaps ques-
tionable. The experimental values of Ct/Cu is used in the computa-
tions. However, the experimental value of Ct /Cu (= 1/8) does not
differ very much from the computed value of C /C (using Eq. (7) t u
= 1/11).
3. The use of Eq, (38) for the determination of the effective
moment of inertia of reinforced beams has been suggested for the
1971 ACI Code (!)(50)~ ). It gives reasonable agreement at loads
very close to the ultimate load also (see Figure 43).
5. 5 Results Reported by Others
The observed load-deflection curves reported by Abeles (56),
Warawaruk, Sozen, and Siess (i:_!_l, Shaikh and Branson (49), and Burns
and Siess (54) are compared with the computed values obtained by
using the methods presented in Sections 5. 1 to 5. 4.
133
Results Reported by Abeles (56)
In his investigation, Abeles reported the load-deflection
response of three groups of rectangular prestressed beams (with
different levels of prestress and steel percentage) under various
conditions of single, repeated, and fatigue load cycles. His primary
interest was in the fatigue loading of pres tressed beams. Single and
repeated load cycle tests were conducted on companion specimens to
obtain a basis of reference. Beams of ordinary and lightweight con
crete were included in the study. Of the 16 beams tested, only A01'~
and ALl~' are used for purposes of this study. Table CS shows the
details of the beams used in this study. Beams AOl~' and ALl'~ were
studied under three cycles of repeated loading. However, no measure
ments of residual deflections were reported. Hence no continuous load
deflection curves under repeated load cycles could be plotted and com
pared with the computed results.
Figure 44 shows the comparison between the computed and
observed values of midspan deflection. On the basis of Figure 44, the
following observations are made:
1. Within the working load (the working load being defined as the
load at which flexural cracking is initiated), the use of the gross sec
tion properties along with the computed modulus of elasticity of con
crete (using Eq. (6)) gives excellent agreement between the computed
(A) Data from Reference (56) ~ 1.8 ,-~--,~~--.~~-.--~~.---=:;.,,.---~--, .0 u Bm AOl* ,.-.~ i:::
"""' i::: 1. 5 ·~
~ 0.43 11
0. 49"
0 • 3 • 6 . 9 1.2 1.5 1.8
(B) Data from Reference (!!) 1.2~~~~~~~~~--,~~~~~~--,
1. 0
6 RB34. 126
e RB34. 093
D RB34. 031 .·
• 08 11
RB34.031 .04" OIE::__....L~-1-~---l~~-'-~-'-~~
0 • 2 • 4 • 6 .8 1.0 1.2
Observedcmidspan deflection in inches
Figure 44 Observed and computed midspan deflection (using Eqs. (38) and (41) for beams under static loading as in (A) (Data from Reference 56) and as in (B) (Data from Reference 41)
135
and observed values of midspan deflection,
2. In the cracked stage, the scatter between the computed and
measured deflections is noticeable. The magnitude of this scatter
increases with an increase in the applied load. This is probably due
to the omission of creep effects in the determination of the effective
moment of inertia using Eq, (38) (see Section 5. 2 for discussion).
However, the magnitude of the scatter is within±_20% for loads which
are about 1. 75 times the working load.
3. In the cracked stage, the computed values (using Eq. (38) for
the determination of effective moment of inertia) of midspan deflection
are greater than the observed values of midspan deflection. Similar
results have been observed by the ACI Committee 435 ('.!_)in the study
of reinforced concrete beams containing 'compression' steel. This
is probably due to the fact that the presence of compressive steel
tends to lower th.e neutral axis and thereby retard the formation of
cracks. (For a discuss ion of this phenomenon as related to other
types of prestressed concrete beams, see Section 5,6),
Results Reported by Warawaruk, Sozen, and Siess (i:.!_)
In a comprehensive study of the strength and behavior in
flexure of pres tressed concrete beams, Warawaruk, et al. reported
the load-deflection response of both post-tensioned and pretensioned
beams. A large number of variables were studied, the most important
136
of which were the steel percentage, type of concrete, loading condi
tions, and type of bonding of reinforcement with the concrete. Of the
82 beams tested, only beams RB34, 126, RB34. 093, and RB34. 031
(pretensioned) are used in this study, The details of these beams are
shown in Table C6. The loading was done statically by a symmetrical
two-point loading system.
Figure 44 shows the comparison between the computed and
observed values of midspan deflection. On the basis of Figure 44,
the following observations are made:
1, Within the working load, the use of the gross section properties
along with the computed modulus of elasticity of concrete gives execl
lent agreement between the computed and observed values of midspan
deflection,,
2. In the cracked stage, the scatter between the computed and
observed values of midspan deflection is noticeable and the magnitude
of this scatter increases with an increase in the applied load. This is
probably due to the omission of creep effects in the determination of
the effective moment of inertia using Eq, (38) (see Section 5. 2 for
discussion). However, the magnitude of the scatter is within:!::_ 20%
for loads which are about 2, 0 times the working load.
3. The beams studies in this report did not have 'compression'
steel. Also, in the cracked stage, the computed values (using Eq,
13 7
(38) for the determination of effective moment of inertia) of midspan
deflection were smaller than the observed values of midspan deflection.
This is consistent with the results described in Section 5. 1 for the
laboratory beams and is probably due to creep effects that have been
neglected in the development of Eq. (38). (For a discussion of this
phenomenon as related to other types of pres tressed concrete beams,
see Section 5. 6.)
Results Reported by Shaikh and Branson (49)
In a comprehensive study of the effects of non-tensioned steel
on the behavior of prestressed concrete beams, Shaikh and Branson
reported the load-deflection response of 12 pretensioned concrete
beams containing various types and quantity of non-tensioned steel.
The details of these beams are shown in Table C7. The loading was
done statically by a symmetrical two-point loading system.
Figure 45 shows the comparison between the computed and
observed values of midspan deflection. On the basis of Figure 45,
the following observations are made:
1. Within the working load, the use of the gross section proper
ties along with the reported modulus of elasticity results in excellent
agreement between the computed and observed values of midspan
deflection.
"' QJ
. .i::: u 1:1
•rl
"' "" •rl s "" QJ .... [
3 (A) Data from Reference (49)
0 Bm I B 1 ti Bm IV B 1 II
eBm I B2 l'lBm IV B2 6BmIB3 181BmIVB3 DBm II Bl
2 •BmIIB2 ®BmIIB3 eBmIIIBl <JBmIIIB2 t>BmIIIB3
deflection under working load for all the beams varied between 0. 2 11
-
s 0 o. 4 11
0 0 u 1 2
(B) Data from Reference (54) .6.---~--~-~--~==r-.--~
ABmJ9
. 5 • Bm J 1 O t----+--+-----11-----i
• BmJll
• 3
l>wl • 2
J9 • 02"
• 1 JlO • 02 11
Jll • 04"
0 3 0 • 1 • 2 • 3 • 4 • 5 • 6
Observed midspan deflection in inches
Figure 45 Observed and computed midspan deflection (using Eqs. (38), (40), and (41) for beams under static loading as in (A) (Data from Reference 49) and for beams under repeated loading as in (B) (Data from Reference 54)
-w 00
139
2. In the cracked stage, the scatter between the computed and
observed values of deflection is noticeable and the magnitude of the
scatter increases with an increase in the applied load. The exclusion
of creep effects in the determination of effective moment of inertia
using Eq. (38) probably causes an underestimation of deflections.
However, the magnitude of the scatter is within ±_20% for loads which
are about 2. 0 times the working load.
3. In the cracked stage, the computed values (using Eq. (38) for
the determination of effective moment of inertia) of midspan deflection
are slightly greater than the observed values of midspan deflection.
This is probably due to the presence of non-tensioned steel which
tends to reduce the creep effect and to further distribute the cracks
along the beam. {For a discussion of this phenomenon as related to
other types of pres tressed concrete beams, see Section 5. 6.)
Results Reported by Burns and Siess {54)
In a detailed study of the effects of repeated loading on the
behavior of reinforced concrete beams, Burns and Siess reported the
load-deflection response of 18 beams. A large number of variables
were studied, the most important of which were the steel percentages,
and the loading regimes. Of the 18 beams tested, only beams J9,
JIO and Jll are included in this study. The beams were unloaded
and reloaded at several stages before and after the yielding of the
140
tension reinforcement. The details of the beams are shown in Table
C8. The loading was done by a symmetrical one-point loading system.
This study indicates that the unloading and reloading from any point up
to the ultimate did not affect the carrying capacity of the beam. The
stiffness of the beam, as measured by the reloading slope of the load
deflection curve was found to depend on the amount of 'inelastic' defor
mation. This is consistent with the results described in Section 5. 2 on
the effects of repeated loading on prestres sed concrete beams.
Figures 45 and 46 show the comparison between the computed
values (using Eq. (40)for the determination of effective moment of in
ertia under repeated loading) and observed values of midspan deflection
under two cycles of loading. The loading stage corresponded to a level
prior to the yielding of the tens ion reinforcement. On the bas is of
Figures 45 and 46, the following observations are made:
1. Within the working load (the working load being defined as the
load at which flexural cracking is initiated), the use of the gross sec
tion properties along with the reported modulus of elasticity of concrete
gives excellent agreement between the computed and observed values of
midspan deflection. The reported and not the computed (using Eq. (6))
modulus of elasticity of concrete because of the large difference that
existed between these two values (a difference of about 20%).
2. At load levels in the cracking range of the beam, the scatter
40
30
"' P< .... "' 20 i::: ....
'O ro 0 10
...:i
0
0
----- Computed
Obs.
=6.60k 7.50k er
k u= 33. 6 35.50k
Beam J9
0.2 0.4 0.6 0.8 0
Comp. k p =3.80
er k
Pu=24. 2
Obs.
3. 9<}<
26 Ok -~""'ield of
Re inf.
Beam J 10
0.2 o.4 o.6 o. 8 Midspan deflection in inches
Observed
Comp.
p =2. 15k er
p =17.25k u
Re inf.
Beam Jll
0 0.2 0.4 0.6 0.8
Figure 46 Comparison of computed and observed values of midspan deflection of beams in Reference (54), under two cycles of repeated loading (three non-composite reinforced beams)
142
begins to be appreciable, and the magnitude of the scatter increases
as the applied load approaches the ultimate load (in this case, the
yielding of the tension reinforcement). However, the scatter is with-
in + 20% for loads which are about 1. 75 times the working load.
Summary of Results Reported by Others
A dimensionless plot (between load and deflection) is also
shown in Figures 47 and 48 for prestressed rectangular and T-beams
(composite or monolithic), respectively. The following observations
are relevant to these figures: (Figures 4 7 and 48)
The allowance of 'severe cracking' in reinforced concrete
beams as compared to 'no cracking' in fully prestressed beams and
'some cracking' (corresponding to the modulus of rupture of concrete)
in partially prestressed beams at service loads, indicates the incon
sistency of the current procedures in the design of reinforced and
prestressed concrete members. One of the reasons for this incon
sistency has been the unavailability of a reliable and simple method
to predict the deflections under 'cracked' conditions for prestressed
concrete members. Figures 4 7 and 48 show the load-deflection
response (on a dimensionless plot) of 24 non-composite prestressed
concrete beams (containing various amounts of tensile, compressive
and non-tensioned reinforcement) and 6 composite prestressed con-
crete beams respectively. Both static and repeated loading resu.lts
are included. Average curves for different steel percentages are
143
1. 6 I 4A01 "11183 ~02 ""'03 '1 ~All X IV81 I <I( El 0
D 181 0 IV82 I +:'.i 1. 5 u • 182 + IV83 I "' Iii I 83 ® A1 ~
j.._ OESIGN__j£ 4-<
~ 1. 4 0 1181 ii A2 • 1182 8 A3 I ZONE I
'1 CJ II 83 11!1 81 <1! p, \711181 111 C1 I I "' 1. 3 '"d Tlll82 <101 I I ..... s AR83412'16R8340931 A ~ 1. 2 ---------> al '" "' Jl 1. 1 I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 (Load, P / Measured ultimate load, P )
um
1. 0
Figure 47 Range of validity of Eqs. (38)(40) and (41) for rectangular beams with different steel pfOrcentages'~ -- included in this dimensionless plot are also the results from studies made on rectangular prestressed beams from References (.!!_), (49), (54) and (~) as well as the current study.
• 2 • 3 .4 • 5 • 6 • 7 • 8 (Load, P /Measures ultimate load, P )
um
• 9 1. 0
Figure 48 Range of validity of Eqs. (38), (40), and (41) for T beams with different steel percentages* -- included in this dimen-11ic>nless plot ate the results from the current study only.
*The values of p and p' in the figure refer to tensile and compressive steel percentages.
145
also indicated in the figures. The computed values of midspan deflec
tion were based on the methods developed in Sections 5. 1 to 5, 4.
For purposes of discussion, the total load range is divided into
three stages --{i) the 'uncracked stage {O - 30% of the ultimate load),
{ii) the 'cracked' stage or "design zone" (30 - 60% of the ultimate load)
and {iii) the 'severely cracked' stage {60 - 100% of the ultimate load).
The following observations refer directly to these figures.
1. In the 'uncracked' stage of non-composite and composite
beams, the variation between the computed and observed values of
midspan deflection is less than±:_20%. The working load of a fully
prestressed beam usually falls within this stage. This confirms the
use of the gross section properties in the determination of midspan
deflections.
2. In the 'cracked' stage of non-composite and composite
prestressed beams, the variation between the computed and observed
values of midspan deflection is still less than+ZO%. However, the ten
dency for this scatter to increase is noticed in the shape of the average
curves. The working load of a partially prestressed beam usually
falls within this range. This suggests the use of the effective section
properties {using Eq. (38) or {40)) as a reasonable method in the
determination of midspan deflections.
146
3. In the 'severely cracked' stage of non-composite and
composite pres tressed beams, the variation between the computed
and observed values of midspan deflection increases markedly as the
applied load approaches the ultimate load. The working load of a pre
stressed beam is, of course, never within this stage. This suggests
the invalidity of the use of Eq. (38) or (40) in the determination of the
effective moment of inertia in this load range.
4. It is noticed that for prestressed beams (in the 'cracked'
and 'severely cracked' stage) and containing only tensile reinforce
ment, the computed values of midspan deflection tend to be smaller
than the observed values of midspan deflection. This appears to be
due to the omission of 'creep effects' in the determination of deflec
tions using the effective moment of inertia (for range of variation, see
(7) below).
5. It is noticed that for prestressed beams (in the
'cracked' and 'severely cracked' stage) and containing both tensile
and compressive reinforcement, the computed values of deflection
tend to be greater than th_e observed values of midspan deflection.
It is believed that this is due to the presence of compressive reinforce
ment which reduces creep and also lowers the neutral axis, thereby
retarding the formation of cracks. Similar observations have been
reported in the AC! Committee report (i) for reinforced concrete
147
beams containing both tensile and compressive reinforcement. (For
range of variation, see (7) below.)
6. It is noticed that for pres tressed beams (in the
'cracked' and 'severely cracked' stage) and containing tensioned and
non-tensioned steel, the computed deflections differ slightly from the
observed values of midspan deflection. However, the variation
between the computed and observed values of midspan deflection for
these beams are small when compared to the variation
between the computed and observed values of midspan deflection for
beams containing only tensioned steel. This is probably due to the
presence of non-tensioned reinforcement that tends to reduce the
creep effect and to further distribute the cracks along the beam.
7. One can safely conclude that 'cracking' (corresponding
to concrete stress es greater than the modulus of rupture) can be
allowed in prestressed concrete members provided the deflections
under such loads satisfy the appropriate serviceability requirements.
When compared to the measured deflections, the use of Eq. (38) for
the effective moment of inertia of prestressed concrete members will
result--(i) in smaller deflections (for prestressed beams containing
only tensile steel), (ii) in larger deflections (for prestressed and rein
forced beams containing both compressive and tensile steel), and (iii)
in very slight deviation from the measured values (for prestressed
148
beams containing both tensioned and non-tensioned steel). However,
the scatter between the computed and observed values of midspan
deflection in all the cases studied herein is within ::+:_20% for loads that
range up to 60-70% of the ultimate load. The corresponding load
range for composite beams is of the order of 75-85% of the ultimate
load.
5. 6 Summary and Conclusions
In Sections 5. 1 and 5. 4 of this chapter, methods were pre
sented for the computation of midspan deflections in both the
1uncracked 1 and the 'cracked' stages of prestressed and reinforced
concrete beams under static or repea"ted loading. Comparisons with
observed values of midspan deflection were made with laboratory
beams of this study (Groups A, B, C, D, E, and F) in Section 5. 1 to
5. 4, and with other data from the literature in Section 5. 5.
On the basis of Figures 32 to 48, and Tables 5, 6, and 7 as
well as the specific conclusions in the earlier sections, the following
general observations are made:
1. In the 1uncracked' or 'elastic' range, the use of the gross sec
tion properties along with the computed values of the elasticity modu
lus of concrete (using Eq. (6)) shows excellent agreement between the
computed and observed values of midspan deflection for both rein
forced and prestressed concrete beams under single or repeated
149
load cycles. (See Sections 5. 1 to 5. 3.)
2. The termination of the 'elastic' or 'uncracked' stage (herein
defined as the cracking load or working load for prestres sed members)
can be predicted with confidence for both reinforced and pres tressed
I concrete beams using the modulus of rupture, fcb· (See Figures
32-43.)
3. The allowance of 'severe cracking' in reinforced concrete
beams as compared to 'no cracking' in fully prestressed beams and
'some cracking' (corresponding to the modulus of rupture of concrete)
in partially pres tressed beams at service loads, indicates the inc on-
sistency of the current procedures in the design of reinforced and pre-
stres!!.ed concrete members. One of the reasons for this inconsistency
has been the unavailability of a reliable and simple method to predict
the deflections under 'cracked' conditions for prestressed concrete
members. Figures 47 and48 show the load-deflection response (on
a dimensionless plot) of 24 non-composite prestressed concrete beams
(containing various amounts of tensile, compressive and non-tensioned
reinforcement) and 6 composite prestressed concrete beams respec-
tively. Both static and repeated loading results are included. Average
curves for different steel percentages are also indicated in the figures.
The computed values of mids pan deflection were based on the methods
developed in Section 5. 1 to 5. 4.
150
For purposes of discussion, the total load range is divided into
three stages -- (i) the 'uncracked' stage (0 - 30% of the ultimate load),
(ii) the 'cracked' stage or 'design zone' (30 - 60% of the ultimate load)
and (iii) the 'severely cracked' stage (60 - 100% of the ultimate load),
The following observations refer directly to these figures.
a. In the 'uncracked' stage of non-composite and composite
beams, the variation between the computed and observed values of mid
span deflection is less than±_20%. The working load of a fully pre
stressed beam usually falls within this stage. This confirms the use
of the gross section properties in the determination of midspan
deflections.
b. In the 'cracked' stage of non-composite and composite
prestressed beams, the variation between the computed and observed
values of midspan deflection is still less than±_20%. However, the
tendency for this scatter to increase is noticed in the shape of the
average curves. The working load of a partially pres tressed beam
usually falls within this range. This suggests the use of the effective
section properties (using Eq, (38) or (40)) as a reasonable method in
the determination of midspan deflections.
c. In the 'severely cracked' stage of non-composite and
composite prestressed beams, the variation between the computed
and observed values of midspan deflection increases markedly as the
151
applied load approaches the ultimate load. The working load of a pre
stressed beam is, of course, never within this stage. This suggests
the invalidity of the use of Eq. (38) or (40) in the determination of the
effective moment of inertia in this load range.
d. It is noticed that for prestressed beams (in the 'cracked'
and 'severely cracked' stage) and containing only tensile reinforcement,
the computed values of midspan deflection tend to be smaller than the
observed values of midspan deflection. This appears to be due to the
omission of 'creep effects' in the determination of deflections using
the effective moment of inertia. (For range of variation see (g) below).
e. It is noticed that for prestressed beams (in the 'cracked'
and 'severely cracked' stage) and containing both tensile and com pres -
sive reinforcement, the computed values of deflection tend to be
greater than the observed values of midspan deflection. It is believed
that this is due to the presence of compressive reinforcement which
reduces creep and also lowers the neutral axis, there by retarding the
formation of cracks. Similar observations have been reported in the
ACI Committee Report ('.!_) for reinforced concrete beams containing
both tensile and compressive reinforcement. (For range of variation,
see (g) below.)
f. It is noticed that for prestressed beams (in the
'cracked' and 'severely cracked' stage) containing tensioned and
152
non-tensioned steel, the computed deflections differ slightly from the
observed values of midspan deflection. However, the variation
between the computed and observed values of midspan deflection for
these beams are small when compared to the variation
between the computed and observed values of midspan deflection for
beams containing only tensioned steel. This is probably due to the
presence of non-tensioned reinforcement that tends to reduce the
creep effect and to further distribute the cracks along the beam.
g. One can safely conclude that 'cracking' (corresponding
to concrete stresses greater than the modulus of rupture) can be
allowed in prestressed concrete members provided the deflections
under such loads satisfy the appropriate serviceability requirements.
When compared to the measured deflections, the use of Eq. (38)
for the effective moment of inertia of prestressed concrete members
will result -- (i) in smaller deflections (for prestressed beams con
taining only tensile steel), (ii) in larger deflections (for pres tressed
and reinforced beams containing both compressive and tensile steel),
and (iii) in very slight deviation from the measured values (for pre
stressed beams containing both tensioned and non-tensioned steel).
However, the scatter between the computed and observed values of
midspan deflection in all the cases studied herein is within:!:_20% for
loads that range up to 60-70% of the ultimate load. The corresponding
load range for composite beams is of the order of 75-85% of the
153
ultimate load.
4. If the concrete and steel stress during a repeated cycle is
below the shakedown limit of concrete (as defined in Section 5. 2)
and the yield strength of steel respectively, the following observations
are valid:
a. The use of Eq. (40) is a reasonable and simple method
of estimating the average effective moment of inertia of pres tressed
and reinforced concrete beams under repeated loading. (See Figures
37-41, 45, 46.) The use of Eq. (40) estimates the recovery during the
unloading cycle. During the unloading cycle, there is no change in
the slope of the load-deflection relationship.
b. Repeated cycles (up to 3 cycles) of loading at a given
load level does not increase the magnitude of the residual deflection
(see Figures 37-39). It is reasonable to expect that further increase
in the number of cycles will not increase the residual deflection any
more.
c. Repeated cycles (up to 3 cycles) of increasing load level
increases the magnitude of residual deflection (see Figures 40-41).
d. The magnitude of the percentage of the total recovery
decreases with increasing load (see sample calculations in Section
5. 2 ).
5. For reinforced concrete beams under 24-hour sustained
154
cracking load, the following observations are valid:
a. The magnitude of the deflection due to sustained load
is a function of the level of the sustained load - - the higher the mag
nitude of the sustained load, the greater will be the deflection under
the sustained load. The use of experimentally determined creep
coefficients predict satisfactorily the deflection under sustained loads
(see Figure 43 ).
b. The use of Eq. (38) is a reasonable and simple means
of estimating the effective moment of inertia of reinforced concrete
beams. The reliability of this equation is confirmed by the fact that
this has been suggested for the 1971 ACI Code (_2_!_).
6. For all the laboratory beams reported in this study, the use
of the equivalent rectangular stress block for concrete gives reason
able agreement between the computed and observed values of ultimate
strength.
7. There was no significant difference either in the strength or
the load-deflection response between composite sections for which
slabs have been cast at different times. This was true under both
single and repeated loading (see Figures 33, 38, 39, 41, 43 and Tables
5, 6, and 7) cycles.
155
Chapter 6
SUMMARY AND CONCLUSIONS
Presented in this study are the results of a comprehensive in
vestigation of non-composite and composite prestressed and reinforced
structures using different weight concretes. Principal emphasis is
placed on the initial plus time-dependent effects (prestress loss, cam
ber, and deflection), and on the load-deflection response under single
and repeated load cycles (with constant as well as increasing -load
levels) into the cracking range.
Systematic design procedures are described for predicting the
material behavior and structural response. Continuous time functions
are provided for all needed parameters, so that the general equations
readily lend themselves to computer solution. Flow charts are explained
and typical computer outputs are given for loss of prestress, camber, and
load-deflection calculations in Appendix F. A summary of general para
meters is also given in Chapter 4 for hand calculations.
These procedures are verified by comparisons between computed
and experimental results for the data of this project, and for additional
data in the literature. These data include normal weight, sand-lightweight,
and all-lightweight concrete, non-composite and composite reinforced
and prestressed members, and both laboratory specimens and actual
156
structures. Ranges of variation are shown and sample calculations are
included for the procedures presented.
The problem, and the objectives and scope of the investigation
are defined in Chapter 1. This chapter also includes a review of liter
ature. A description of the experimental investigation of this project is
given in Chapter 2.
Systematic procedures are described in Chapter 3 for predicting
strength and elastic properties, creep and shrinkage characteristics of
different weight concretes, types of curing, and types of cement (Eqs.
2 - 13). Standard equations and correction equations for significant
conditions other than "standard" are outlined for design purposes. This
chapter was developed in this project(~) and in Reference (..!_~). Com
parisons between experimental and computed results are shown to be
quite satisfactory for the data of this project (Figures 2 - 7 and B3).
Procedures for predicting the initial plus time-dependent loss of
prestress and camber of prestressed beams and deflection of reinforced
beams are presented in Chapter 4 (Eqs. 14 - 34). Computed results by
these equations, using both experimental material parameters and gen
eral or average parameter.s, are compared with experimental results
for the laboratory beams and sand-lightweight composite bridge of this
project; and with additional data in the literature (Figures 8 - 29 and
Tables I - 4). Separate steel relaxation tests were conducted, and the
contribution of steel relaxation to loss of prestress in beams (as distin
guished from relaxation tests at constant length) is included in a
157
rational manner. It is concluded that the results in Chapter 4 serve to
substantiate the prediction methods described. The approximate equations
may be used for rough calculations only in some cases.
The ultimate loss of prestress for the sand-lightweight concrete
(composite) prestressed bridge girders was 29% to 31% (see Figure 1,
and Tables 1 and 3). It was determined that loss percentages for bridges
under similar conditions using normal weight concrete will normally be
of the order of 25%; and using all-lightweight concrete will normally be
of the order of 35% or higher. Higher losses for the lighter concretes,
for example, are due primarily to the lower modulus of elasticity
(higher elastic strains for a given stress level), and not, necessarily,
to greater creep and shrinkage behavior.
With respect to different slab casting schedules for composite
prestressed and reinforced beams, an earlier slab tends to reduce the
creep curvature by forming an earlier composite section, and also by
reducing differential shrinkage. On the other hand, the creep effect
for the precast beam concrete under the earlier slab loading tends to be
greater. It appears from this study that the net result of these offsetting
effects is beneficial in both prestressed and reinforced beams (earlier slab
reduces pres tress loss, camber, and deflection). It was found in this
study that the beneficial effect of an earlier slab (3 to 4 weeks versus
9 to 10 weeks herein) is relatively small for prestressed beams and
relatively significant in reinforced beams. The decrease in computed
ultimate prestres s loss and camber for the laboratory beams and bridge
girders herein (see Figures 1, ll, 12, 15, 16, 18, and Tables 1, 2)
was negligible for the laboratory beams; and 2% less prestress loss,
158
and 0.10" less midspan camber, for the bridge girders. Only the numer
ical camber, and not the percentage, is meaningful for the bridge gir
ders, because the total camber is near zero due to the heavy deck slab.
The decrease in the ultimate deflection of the laboratory composite beams
was 0.13" or 30% (see Figures 1, 17, and Table 2). The reason for the
difference in the relative effects between prestressed and reinforced
beams has to do with the offsetting effects of prestress and dead load
(including slab dead load) in the one case, and only additive dead load
effects in the case of reinforced beams.
A detailed discussion of the experimental results and conclusions
is also given in Chapter 4.
From the results in Chapter 4, it is concluded that the procedures
presented will normally agree with actual results within _:'.:15% when
using experimentally determined material parameters. The use of the
general or average material parameters herein predicted results that
agreed with actual results in the range of..± 30%. With some knowledge
of the time-dependent behavior of concretes using local aggregates and
under local conditions, it is concluded that one should normally be able
to predict initial plus time-dependent loss of prestress, camber, and
deflection within about _:'.:20%, using these procedures. Some 41 lab
oratory specimens and actual structures were included in Chapter 4.
In the cases compared, it is noted that most of the results are consider
ably better than these limits.
This project is thought to be the first such comprehensive study
of the initial plus time-dependent material behavior and related
159
structural response of both non-composite and composite structures
using different weight concretes.
Developed in Chapter 5 for the first time is a simple and efficient
design method for predicting the entire short-time load-deflection curve
(or a single point, such as at maximum load) under repeated load cycles
into the cracking range for both prestressed and reinforced members.
This method is based on a procedure developed by Branson (50), (.'.!_),
(30), (42) for predicting the deflection of reinforced beams under single
cycle loading and adopted for the 1971 ACI Building Code (22:_), and applied
to prestressed beams under single-cycle loading by Shaikh and Branson
(49). The effects of increasing load levels in subsequent cycles, and of
24-hour sustained loading are also included. Eqs. (35) - (41), the accom
panying descriptions, Figures 32 - 43, Tables 5 - 7, and the correspond
ing sample calculations serve to illustrate these procedures.
The reliability of the procedures described are indicated by com
parisons between computed results and the experimental data of this
project, and with data in the literature (Figures 31 - 34, 37 - 48, and
Tables 5 - 7).
It was found (Figures 37 - 42, and Table 6) that repeated load
cycles (up to 3 cycles in this project) of short duration did not increase
the deflection at a given load level nor the residual deflection after un
loading. However, repeated cycles to increasing load levels did in
crease the residual deflection after unloading, and also increased the
magnitude of the deflection at a given load level when reloaded (Figures
160
40 - 42 and Table 6). Similar results have been shown in Reference
(54). This is attributed to the effect of greater crack development at the
higher loads, and correspondingly greater residual crack effects.
A detailed discussion of the experimental results and conclusions
is also given in Chapter 5.
From the results in Chapter 5, it is concluded that the procedures
presented for predicting load-deflection behavior of reinforced and
prestressed members will normally agree with actual results within
+ 20% for loads as high as 60% to 70% of the ultimate load for non-
composite beams and as high as 75% to 85% for composite beams under
both single and repeated load cycles. This included partially pre stressed
beams l<11aded well into the cracking range. The accuracy is generally
better than~ 20% for normal working load levels. Some 38 non-composite
and composite specimens were included in Chapter 5 (Figures 31 - 34,
37 - 48, and Tables 5 - 7).
With the aid of the material parameter equations presented in
Chapter 3, and the procedures developed in Chapters 4 and 5, the
structural designer can more reliably than in the past predict the initial
plus time-dependent prestress loss, camber, and deflection (including
effects of repeated load cycles) of non-composite and composite rein
forced and prestressed structures of different weight concretes. As a
result of this study, he can also make a better judgement as to the
reliability of his computational procedures and the range of variation to
be expected between computed and actual results, depending primarily
on the degree of care with which the material properties and parameters
(mainly creep and shrinkage) are determined for a given design.
161
LIBT OF REFERENCES
1. Davis, R., "A Summary of Investigation of Volume Changes in Cement Mortars and Concrete Produced by Causes Other Than Stress," ASTM, Proceedings, V. 30, Part I, 1930, pp. 668-685.
2. Carlson, R. W., "Drying Shrinkage of Concrete as Affected by Many Factors, 11 ASTM, Proceedings, V. 38, Part II, 1938, pp. 419-440.
3. Hveem, F. N., and Tremper, B., "Some Factors Influencing the Shrinkage of Concrete," ACI Journal, Proceedings, V. 53, No. 8, Feb. 1957, pp. 781-802.
4. ACI Committee 435, "Deflections of Reinforced Concrete Flex-ural Members, " ACI Journal, Proceedings, V. 63, No. 6, June 1966, pp. 637-674.
5. Lorman, W. R., "The Theory of Concrete Creep," ASTM, Proceedings, V. 40, 1940, pp. 1082-1102.
6. McHenry, Douglas, "A New Aspect of Creep in Concrete and Its Application to Design," ASTM, Proceedings, V. 43, 1943, pp. 1969-1984.
7. Neville, A, M., "Theories of Creep in Concrete," ACI Journal, Proceedings, V. 52, No. 1, Sept. 1955, pp. 47-60.
8. Ross, A. D., "Creep of Concrete Under Variable Stress," ACI Journal, Proceedings, V. 54, No. 9, Mar. 1958, pp. 739-758.
9. Troxell, G. E.; Raphael, J. M.; and Davis, R. E., "Long Time Creep and Shrinkage Tests of Plain and Reinforced Concrete," ASTM, Proceedings, V. 58, 1958, pp. 1-20.
162
10. Kesler, C. E., and Ali, I., "Mechanisms of Creep, 11 Symposium on Creep of Concrete," ACI Special Publication No. 9, 1964, pp. 35-63.
11. Meyers, B. L.; Slate, F. O.; and Winter, G., "Time-Depen-dent Deformation and Microcracking of Plain Concrete, 11
ACI Journal, Proceedings, V. 66, No. 1, Jan. 1969, pp.
12. Meyers, B. L., and Neville, A. M., "Creep of Concrete: Influencing Factors and Prediction," Symposium on Creep of Concrete, ACI Special Publication No. 9, 1964, pp. 1-33.
13. Pauw, A., and Chai, J. W., "Creep and Creep Recovery for Plain Concrete," Missouri Cooperative Highway Research Programme, Report No. 67-8.
14. Hansen, T. C., and Mattock, A. H., "Influence of Size and Shape of Member on Shrinkage and Creep of Concrete, 11
ACI Journal, Proceedings, V. 63, No. 2, Feb. 1966, pp. 267-290.
15. Jones, T. R.; Hirsch, T. J.; and Stephenson, H. K., "The Physical Properties of Stt"uctural Quality Lightweight AggI"egate Concrete, 11 Texas Transportation Institute, Texas A & M University, College Station, Texas, August 1959, pp. 1-46.
16. ACI Committee 213, 'Guide for Structural Lightweight Aggre-gate Concrete," ACI Journal, Proceedings, V. 64, No. 8, Aug. 1967, pp. 433-470.
17. Pfeifer, D. W., "Sand Replacement in Structural Lightweight Concrete- ~Creep and Shrinkage Studies, "ACI Journal, Proceedings, V. 65, No. 2, Feb. 1968, pp. 131-142.
18. Christiason, M. L., "Time-Dependent Concrete Properties Related to Design--Strength and Elastic Properties, Creep and Shrinkage," MS Thesis, University of Iowa, Iowa City, Feb. 1970.
163
19. Schumann, C, G., "Creep and Shrinkage Properties of Light-weight Aggregate Concrete Used in the State of Iowa, " MS Thesis, University of Iowa, Iowa City, Jan. 1970.
20. Finsterwalder, Ulrick, "Ergenbnisse von Kriech und Schwind-messungen an Spannbetonbauwerken, 11 Beton und Stahlbetonbau (Berlin}, V. 53, No. 5, May 1958, pp. 136-144.
21. Lofroos, W. N., and Ozell, A. M., "The Apparent Modulus of Elasticity of Prestressed Concrete Beams Under Different Stress Levels, " Pres tressed Concrete Institute Journal, V. 4, No. 2, Sept. 1959, pp. 23-47.
22. Mattock, Alan H., "Precast-Prestressed Concrete Bridges; 5. Creep and Shrinkage Studies," Journal, Research and Development Laboratories, Portland Cement Association, V. 3, No. 2, May 1961, pp. 32-66.
23. Branson, D. E,, and Ozell, A. M., "Camber in Pres tressed Concrete Beams," ACI Journal, Proceedings, V. 57, No. 12, June 1961, pp. 1549-1574.
24. Corley, W. G,; Sozen, M. A.; and Siess, C. P., "Time-De-pendent Deflections of Prestressed Concrete Beams," Bulletin No. 307, Highway Research Board, 1961, pp. 1-25.
25. Branson, D. E,, "Time-Dependent Effects in Composite Con-crete Beams," ACI Journal, Proceedings, V. 61, No. 2, Feb. 1964, pp. 213-230.
26. Zia, P,, and Stevenson, J. F., "Creep of Concrete Under Non-Uniform Stress Distribution and Its Effect on Camber of Prestressed Concrete Beams," Report of Highway Research Programme No. ERD-110-R, JUne 1964, pp. 1-110.
27. Sinno, R., "The Time-Dependent Deflections of Prestressed Concrete Bridge Girders," Dissertation, Texas A & M University, 1968.
28. Yang, D. D., "Creep in Prestressed Lightweight Aggregate Concrete," Dissertation, Texas A & M University, 1966,
164
29. Scordelis, A. C., Subcommittee Chairman, Branson, D. E., and Sozen, M. A., ''Deflections of Pres tressed Concrete Members, 11 A.CI Committee 435, Subcommittee 5 Report, A.CI Journal, Proceedings, V. 60, No. 12, Dec. 1963, pp. 1697-1728.
30. Branson, D. E., "Design Procedures for Computing Deflec-tions," A.CI Journal, Proceedings, V. 65, No. 9, Sept. 1968, pp. 730-742.
31. Pauw, Adrian, and Breen, J. E., "Field Testing of Two Pre-stressed Concrete Girders," Highway Research Board Bulletin 307, pp. 42-63, 1961.
32. Young, J. A., "Field Observation of Five Lightweight Aggre-gate Pretensioned Prestressed Concrete Bridge Beams, 11
Final Report, Iowa Highway Research Board Project No. HR-104, pp. 1-39, 1969.
33. Branson, D. E.; Meyers, B. L.; and Kripanarayanan, K. M., "Time-Dependent Deformation of Non-Composite and Composite Pres tressed Concrete Structures, " Iowa Highway Commission Research Report 69-1, Feb. 1969, pp. 1-80. Also condensed papers presented at the 49th Annual Meeting, Highway Research Board, Washington, D. C., Jin. 1970, pp. 1-42; and at the 6th Congress, Federation Internationale de la Precontrainte, Prague, Czechoslovakia, June 1970, pp. 1-28.
35. Abeles, P. W., "Static and Fatigue Tests on Partially Pre-stressed Concrete Constructions, 11 A.CI Journal, .!3:£: ceedings, V. 50, No. 7, Dec. 1954, pp. 361-376.
36. Abeles, P. W., "Partial Pres tressing and Possibilities for Its Practical Application, 11 Prestressed Concrete Institute Journal, V. 4, No. 1, June 1959, pp. 35-51.
37. Abeles, P. W., "Partial Prestressing in England, 11 Prestressed Concrete Institute Journal, V. 8, No. 1, Feb. 1963, pp. 51-72.
165
38. Abeles, P. W., "Studies of Crack Widths and Deformation Under Sustained and Fatigue Loading," Prestressed Concrete Institute Journal, V. 10, No. 6, Dec. 1965.
39. Burns, N. H., "Moment Curvature Relationships for Partially Prestressed Concrete Beams," Prestressed Concrete Institute Journal, V. 9, No. 1, 1964, pp. 52-63.
40. Hutton, S. G., and Loov, R. E., "Flexural Behavior of Pre-stressed, Partially Prestressed and Reinforced Concrete Beams, "ACI Journal, Proceedings, V. 63, No. 12, Dec. 1966, pp. 1401-1408.
41. Warawaruk, J., Sozen, M. A., and Siess, C. P,, "Strength and Behavior in Flexure of Prestressed Concrete Beams," Engineering Experiment Station Bulletin No. 464, University of Illinois, August 1962, pp. 1-105.
42. Branson, D. E., Subcommittee Chairman, "Prediction of Creep, Shrinkage and Temperature Effects in Concrete Structures," Subcommittee II, ACI Committee 209, Draft Report, April 197 0, pp. 1- 32.
43. Keeton, J. R., "Study of Creep in Concrete, Phases 1-5," Technical Reports Nos. R333-I, II, III, U.S. Naval C. E. Lab., Port Hueneme, Calif., 1965.
44. The California Producers Committee on Volume Change and Affiliated Technical Organizations, "Drying Shrinkage of Concrete," p. 1-40, Mar. 1966.
45. Magura, D. D.; Sozen, M. A.; and Siess, C. P., "A Study of Relaxation in Prestressing Reinforcement," Prestressed Concrete Institute Journal, V. 9, No. 2, Apr. 1964, pp. 13-58.
46. Antill, J. M., "Relaxation Characteristics of Prestressing Tendons," Civil Engineering Transactions, Inst. of Engr., Australia, V. CE 7, No. 2, 1965.
47. Evans, R. H., and Bennet, E. W., Prestressed Concrete, Wiley, New York, 1958.
166
48. Rogers, G. L., "Validity of Certain Assumptions in the Mechanics of Prestressed Concrete, " ACI Journal, Proceedings, V. 49, No. 7, Dec. 1953, pp. 317-330.
49. Shaikh, A. F., and Branson, D. E., "Non-Tensioned Steel in Prestressed Concrete Beams," Prestressed Concrete Institute Journal, V. 15, No. 1, Feb. 1970.
50. Branson, Dan E., "Instantaneous and Time-Dependent Deflec-tions of Simple and Continuous Reinforced Concrete Beams," Part I, Report No. 7, Alabama Highway Research Report, Bureau of Public Roads, Aug. 1963, (1965), pp. 1-78.
51. ACI Committee 318, "Proposed Revisions to the ACI-318-63 Code," ACI Journal, Proceedings, V. 67, No. 2, Feb. 1970, pp. 77-186.
52. Noble, P. M., "The Effect of Aggregate and Other Variables on the Elastic Properties of Concrete," Proceedings, ASTM, V. 31, Part I, 1931, pp. 399-426.
53. Shah, S. P., and Winter, G., "Response of Concrete to Repeated Loads," RILEM International Symposium on the Effects of Repeated Loading on Materials and Structural :E:lements, Sept. 1966, Mexico.
54. Burns, N. H., and Siess, C. P., "Repeated and Reversed Loading in Reinforced Concrete, " ASCE Journal (Structural Division), V. 92, Paper No. 4932, Oct. 1966.
55. Ruiz, M. W., "Effect of Repeated Loads on the Rotation Capacity of Reinforced Concrete Beams," Dissertation, Cornell University, Sept. 1968.
56. Abeles, P. W., Brown, E. I., and Woods, J. 0., "Report on Static and Sustained Loading Test, " Prestressed Concrete Institute Journal, V. 13, No. 4, Aug. 1968, pp. 12-32.
57. Reichart, T. W., "Creep and Drying Shrinkage of Lightweight and Normal-Weight Concretes, " NBS Nomograph 74, National Bureau of Standards, Mar. 1964.
58. Shideler, J. J., "Lightweight Aggregate Concrete for Structural Use," ACI Journal, Proceedings, V. 54, No. 4, pp. 299-328, Oct. 1957.
Ap 1
APPENDIX A
Appendix A includes the details of the laboratory speci
mens and the bridge girders as well as the different
types of concretes used in this project. This also
includes details of the creep and shrinkage specimens
such as the age of loading, ambient relative humidity,
etc.
TABLE Al
. ' ' DETAILS OF LABORATORY BEAMS (GRPS A B C) AND BRIDGE GIRDERS L=86 ',
aAll Beams are 6 11 x 8 11, d=6 11
, Span= 15", bslabs are 20" x 2" 7 11 slab
Beam Groun Group A Group B Group C Bridge
Ream No. Al A2 A3 Bl B2 B3 Cl CZ C3 152-156
c f
Beam D CJ CJ D u u [J u 1f TI Eccentricity in 2.00 2.00 2.00 2.00 2.00 2.00 2,00 2.00 2.00
14.50 6.20
Prestressing . 2-3/8 3-5/16 1-3/8 3-5/16 3-5/16 3-5/16 2-3/8 2-3/8 2-3/8 30-172 ~trand dia in 1-5/16 1-5/16 1-5/16 1-5/16 1-5/16
eA s in2 0.2176 o. 17 34 o. 1377 o. 1734 o. 1734 o. 1734 0.2176 0.2176 0.2176 4.56
Measured stress in all strands Sand Lt. Wt. concrete.
TABLE A 1 (Cont'd)
Measured s tress of bridge girders
in all strands of lab beams = ( 1 72±_ 4) ks i. = 190 ksi. All beams are made of ldealite -
b Six gage WWF, 6" x 6", (As = O. 058 in2 /ft width), slab steel placed in center of slab. No. 3 CT-Stirrups in form of ties for composite slab are spaced at 6 11 c/c in end quarter span and at 22 -1/2" cc in middle half of beam.
c
d
e
Strands placed so that lateral eccentricity is eliminated.
These stresses are computed using the Measured F., t= top fiber stress, b= bottom fiber stress. 1
These initial stresses refer to prestressed section in all cases. The stresses in the case of laboratory beams refer to the end section only. The rectangular (6" x 8") beam dead load, extreme fiber stress at midspan= 218 psi.
The ultimate strength and yield strength (0. 1% offset) were: for the laboratory beam steel 250 ksi and 235 ksi, respectively, and for'the bridge girder steel 270 ksi and 250 ksi, respectively.
f The lower values in this column refer to the center of the girder.
TABLE A2
DETAILS OF LABORATORY BEAMS (GRPS. D, E AND F)
aAll Beams are 6" x 8", d=6", Snan= 15', bslabs are 20" x 3" Beam Group Group D Group E c Group F
Beam No. Dl D2 D3 El E2 E3 Fl F2 F3
Beam [] [] [] [J w w IJ 1d l:J Eccentricity in 1.75 2.00 2.00 1. 75 1. 75 1. 75 2.00 2.00 2.00
Prestressing . 4-3/8 4-5/16 1-1/4 4-3/8 4-3/8 4-3/8 3-1/2 3-172 3-1/2 Strand dia m 3-5/16
e . 2 As 1n o. 3196 0.2312 0.2090 0.3196 0.3196 0.3196 0.6000 o. 6000 0.6000
d Concrete t=+385 t=+42 l t=+369 t=+375 t=+ 370 t=+370 Stresses at release of b=-2585 B=-2049 b=-1831 b=-2585 b=-2591 b=-2600 pres tress, psi
For footnotes see following page.
TABLE A2 (Cont'd)
a • 3/8" Strand, o 5/16" Strand, a 1/4" Strand, x 1/2" bar, Measured stress in all strands of lab beams = (175 ±.. 2) ksi.
b See Footnote b, Table A 1
c The value of p for reinforced beams is As/bd.
d These stresses are computed using the Measured Fi: t =top fiber stress, b =bottom fiber stress. These initial stresses refer to the prestressed section in all cases. The stress in the case of laboratory beams refer to the end section only. The rectangular (6" x 8") beam dead load, extreme fiber stress at midspan are 178 psi, 208 psi, 208 psi for the beams of Group D, E, and F, respectively.
e See Footnote e, Table A 1
TABLE A3
DETAILS OF CONCRETE MIXES AND MIXING PROCEDURE FOR LT-WT CONCRETES
Cone rete for Description Grps A, B, C& Group D Group E
Bridge Girders Mix design objectives
Cone. Qty, 1 cu vd 1 cu vd 1 cu vd Cone. str. @28d 5000 psi 5000 psi 5000 psi
F. aggregate Sand - 1395 lbs Haydite agg. (3/16" Sand - 1150 lbs to dust) - 950 lbs
c. aggregate ldealite Agg. (60% Haydite Agg. (3/ 4" Haydite Agg. (3/4" of 3/4 to 5/16 & to 11'4) - 700 lbs to"' 4) - 825 lbs 40% of 5/16 to 4'"8)
822 lbs
Water 35.0 gal 42. 0 gal 42. 0 gal
Dar ex 6. 5 oz 7. 0 oz 6. 5 oz
WRDA 50 oz 53, 5 oz 50 oz
Mixing procedure: 1, Proportion and batch fine aggregate and coarse aggregate. 2, Add 50% of total water requirement. 3. Mix for approximately 2 min. 4. Proportion and batch cement. 5, Add 12. 5% of water requirement.
Group F
1 cu vd 4000 psi
611 lbs
Sand - 1250 lbs
Haydite Agg. (3/4" to #4) - 825 lbs
40. 0 e:al
5. 7 oz
43. 5 oz
6. Add Darex (in solution with 3 gallons of water), WRDA and the remaining water while adjusting to 2-1/2" slump.
TABLE A4
a-gCONCRETE PROPER TIES (GRPS A, B, C AND BRIDGE GIRDERS), TEMPERATURE AND HUMIDITY DA TA
Concrete Batch Gp.A Gp. B Gp. C Slab Slab Slab Slab 1 Bridge
Property SLt. Wt SLt. Wt SLt. Wt BZ CZ B3 C3 Lt. Wt N. Wt N. Wt N. Wt N. Wt
of 3.04 Modulus - - - - a. 3. 20 - - - - - - - - a.
Elasticity ps~ -- -- b. 3.33 - - - - -- - - b. 3. 10
at 7 days x 10 3,68 ~ c. 3,55 - - -- - - - - c. 3. 32
c 3. 28 Modulus of -- -- a. - - - - -- - - --
Elasticity psi - - - - b. 3,58 - - -- - - - - --at 28 days x 106 ~ ~ c. ~ ~ 3.97 !:..!.!. ±:.-22. 3,47
For footnotes, see following page.
!;;Bridge Slab
N. Wt
3500
- -
145
- -
--
- -- -- -
- -- -
3,41
TABLE A4 (Cont'd)
a Lab. temp: 61-85 deg. F., avg. temp. 78 deg. F. Lab. relative humidity: 2.5-61%, avg. rel. hum. 40%. Avg. rel. hum. for central Iowa (from U.S. Weather Bur.): Jan. -79%, July-66%, Mean Annual 71%. For Spr-Sum-Fall, use 70%.
b Stress levels for creep tests were approx. design stresses for lab. beams: I I
Mix Strength, fc, at 7 days Stress Level for Creep Tests % of 7d - fc
Gp. A 6700 psi 2.010 psi 30% Gp. B 5500 1375 2.5 Gp. C 6150 1845 30
C I The modulus of elasticity values are as follows: a. Measured secant (to O. 5 fc) mod. of el.,
d
b. Measured initial tangent mod. of el., c. All values underlined are computed using Ee = 33 J;,;3£~ , psi.
Computed values Girder No.
152. 153 154 155 156
of modulus of elasticity at release for bridge Age at Release Strength at Rel.
2. days 5160 psi 2. 4670 2. 4685 3 5130 3 4440
girders: cMod. of El. at Rel.
3. 19 x 106 psi
.h.Q.! ~ 2.:12. 2..96
e Computed mod. of el. of pres. units at time of slab casting, cEc x 106 psi: Gp. B-·-~· 4. 30; Gp. C--4.2.3, 4.44; Girders 152., 153, 154--3.50; Girders 155, 156--~.
f Concrete specimens for data in this column obtained from casting yard for Bridge Girders 155 and 156. Measurements made in laboratory.
g "Design" values were used for bridge slab concrete.
Property
b I
fc (Release) psi
I
fc (28 days) psi
Unit Wt (Wet pcf
U. Wt (Dry-7d) pcf
Meas. Air Ent, %
Slump in
cModulus of psi
Elasticity 106 at Release x
cModulus of psi
Ehsticity at 28 Days 106 x
TABLE A5
aCONCRETE PROPERTIES (GRPS D, E, & F), TEMPERATURE AND HUMIDITY DATA
Concrete Batch Slab
Gp D Gp E Gp F E2 A. Lt. Wt S. Lt. Wt. S. Lt. Wt N.Wt
Stress levels for creep tests were approximate design stresses for lab. beams:
Mix Gp D Gp E Gp F
Age @ release 7 days 9 days
21 days
Strength@ release 4150 psi 4250 psi 3650 psi
Stress level for % initial creep tests
2000 psi 2000 psi 1000 psi
stress 48% 47% 27%
The age at release for Gps D and E refer to the age at release of prestress and for Gp F this refers to the age at which the reinforced beams were in position.
c 131 All values are computed using E = 33 ,./w- fc , psi.
c
..... 0
Ap 11
TABLE A6
CONCRETE PROPER TIES OF LAB BEAMS AT "LOAD-DEF" STUDIES
Computed using Eq. (2). The beams of Group A, B, C, D, E and F were aged 367, 187, 187 .• 187, 189 and 189 days respectively at the time of the load deflection studies.
For lightweight concrete in a drying condition, the modulus of rup-ture ranges from 5Jf';; to llffc. The observed values of the modulus of rupture correspond to approximately 6 ]fr .
Obtained by bending tests on plain concrete members.
The concrete strength of slab concretes of B2, B3, C2, C3, E2, E3, F2, and F3 were 5500, 5760, 4720, 4860, 4200, 4860, 4860 and 4750 psi, respectively.
Ap 12
APPENDIX B
Appendix B includes a discussion of the variables that
affect creep and shrinkage of concretes as well as a
discussion of the correction factors for these variables
with relation to the method developed in the text.
Ap 13
APPENDIX B
Discussion of variables affecting creep and shrinkage (i_)(Q)(~)(42)(~)
Concrete undergoes time-dependent deformations under the
action of sustained loads that are substantially greater than those of
a corresponding unstressed specimen. These additional strains due
to the effect of sustained stress are attributed to creep of the concrete.
Current nomenclature regarding creep of concrete is summarized in
Figure B 1,
When specimens are subjected to uniform axial stress, only
normal strains (both elastic and inelastic) are usually considered.
The elastic strains are stress dependent and recoverable. These
strains include both time-independent and time-dependent strains.
The time-independent elastic strain is also referred to as initial or
instantaneous strain.
The stress independent component of the inelastic strain is
normally called shrinkage. This strain is partially reversible. The
s tress dependent irrecoverable strains include microcracking effects
as well as shrinkage or drying creep resulting from moisture migra
tion due to applied stress. The drying creep cannot be separated
from the irreversible shrinkage.
The total creep strain consists of (a) Basic creep--delayed
strain due to the interaction between solid and fluid phase, (b) Drying
I Elastic Strains I I Inelastic Strains I ' ' • .L • I
and time init. shrink- 14. Stress Distr. age considered 15. Stress magnitude
11. Duration of load period 16. Stress rate
9 10 11 12 13
Loading History
14 15 16
Stress Condition
Environmental Conditions
Loading Conditions
------4-.I Time-Dependent Strain Variables
Figure B2. Time-Dependent Strain Variables
~' Parameters studied by Jones (__!2), and used in this report + These numbers refer to the parameters listed above
Ap 17
1, 1 Minimum thickness of member
1. 2 Water-cement ratio in the form of slump and cement content
1, 3 Mix proportions in the form of percent fines and air content
1, 4 Environmental humidity
1, 5 Time of initial loading and time initial shrinkage considered
Presented here is a summary of the principal variables that
affect creep and shrinkage (!.Q), (g), (13), (!.2_), (.!_~)in most cases.
The corresponding nominal correction factors, based on the standard
conditions herein, are given in the text and in Figure B3 (13), (15), - -(!_~). The results in Figure B3, and equations for these curves, were
developed in Reference (.!.§_).
The following comments refer to the nominal correction factors
for creep and shrinkage (from Figure B3 ), which are normally not
excessive and tend to offset each other. For design purposes in most
cases, these (except possibly for the effect of member size as dis-
cussed in the text) may normally be neglected:
Creep correction factors
Slump: C, F. = O. 95 for 2", 1. 00 for 2. 7", 1. 02 for 3", 1. 09 for 4", 1. 16 for 5". Comment--Tends to be offset by effect of member thickness. May be marginal but normally can be neglected.
Cement content (sacks/cu.yd.): C. F. = 1, 00. No correction factor required for concrete of say 5 to 8 sacks per cu. yd. at least.
Percent fines (by wt. ) : C. F. = O. 95 for 3 Oo/o, 1. 00 for 5 0%, 1, 05 for
70%. Comment--Normally negligible.
001.2.-----.--.----71 '" 0 ... u .!'11.01-----±-7""-'-=::J_---i
p., ~ <1J 0 ~ ·.;::: A. Lt, M u u 0. 8(--U-.olL--'--------1
~ 0(17),I,S.Lt,M ~ 57 III, A. Lt, S uO. 6 "57 III N. Wt S
0 2 4 6 a. Slump (in)
0.8
O (15), I, A. Lt, M L:,.(58), I, A. Lt, M
4 6 8 10 b. Cement content
(sacks/cu.yd.)
o(.!i), I, A. Lt,
0 (15), I, A. Lt, 0.8~------~
20 40 60 80 c. Percent fines by
wt. (<1/4 sieve)
o.
4 8 12 16 d. Air content (%)
0 6(58), I, A. Lt,
.8'--.=::...0."-'--'----'--' 6 0(15), I, A. Lt, o. .......-~~---~
jW,5 , I, A. Lt, M
TI , I, S. Lt, M , III, A. Lt,
0 1 , III, N. Wt, • .-.=:.'-"'C.:....-'-......;.-'"'-'
4 6 8 10 40 60 80 f. Cement content
(sacks/cu.yd.)
20 g. Percent fines by
wt. (< 114 sieve)
4 8 12 16 h. Air content (%)
'" 0 ... u
1. 2
<IJ.!'10.9 "" nj ~
.>: 0 ~ .....
•C!.!l· !eshl3ood °C!_!), (E:shh300d
·~ 0 o. 6 - Use for fi5 ~ ,.. 1 yr.
0 o 3 dr in u • '""'"'W'-'~----.....:l 6 12 18 24
Figure B3.
j. Minimum thickness (in)
Nominal creep and shrinkage correction factors for the parameters shown from Ref. 18. Notation: I, III -- type cement; N. wt., S. Lt., A. Lt. Weight concrete; M, S Moist, Steam curing
Ap 19
Air content (in%): C. F. = 1. 00 up to 6%, 1. 09 for 7%, 1. 17 for 8%. Comment--Tends to be offset by effect of member thickness. May be neglected for say up to 7% air.
Minimum thickness of member: C.F. = 1.00 for 6" or less, 0.82 for 12". Comment--Tends to be offset by effect of slumps greater than 3" and air contents greater than 6%. Can normally be neglected for members up to about 10" to 12".
Shrinkage correction factors
Slump: C. F. ::: O. 97 for 2 11, I. 00 for 2. 7 11
, 1. 01 for 3 11, 1. 05 for 4",
1. 09 for 5". Comment--Tends to be offset by effect of member thickness. Normally can be neglected.
Cement content (sacks/cu. yd.): C. F. = O. 87 for 4 sacks, 0. 95 for 6 sacks, 1. 00 for 7. 5 sacks, 1. 09 for 10 sacks. Comment- -Normally negligible for say 5 to 8 sacks per cu. yd. at least.
Percent fines (by wet.): C. F. = O. 86 for 40%, 1. 00 for 50%, 1. 04 for 70%. Comment--May be marginal but normally can be neglected.
Air content (in%): C. F. = O. 98 for 4%, 1. 00 for 6%, 1. 03 for 10%. Comment- -Normally negligible.
Minimum thickness of member: C. F. = 1. 00 for 6" or less, O. 84 for 9". Comment--Tends to be offset by effect of slumps greater than 3". Can normally be neglected for members up to about 8" to 9" minimum thickness.
Ap 20
APPENDIX C
Appendix C includes the details of the test beams from References
(~), (11), (f.1), and (2!_). The loss of prestress and camber of these
beams have been discussed in the text on the basis of the methods
developed therein. Also included are the details of the test beams
from References (Q), (i.2.J, (54), and (56). The load-deflection
response of these beams have been discussed in the text on the basis
of the methods developed therein.
TABLE Cl PROPERTIES OF TEST BEAMS AT UNIVERSITY OF FLORIDA (23)
Cable a I
b Cone. Beam Eccentricitv Concrete Curing Loading A Fa fc rel Stress Profile End Center Type Cond. Age in~ (kip) (psi) End (psi) Center (psi)
1 STRT 2. 16 2.31 Nr wt MC 28d 1. 32 57.9 5030 902 805
2 STRT 2. 02 2. 16 Nr wt MC 28d 1.32 65. 1 5030 982 891
3 STRT 2.20 2.44 Nr wt MC 28d 1. 32 101. 5 5030 1602 1568
4 STRT 1. 91 2.26 Nr wt MC 28d 1. 32 99.9 5030 1457 1460
5 STRT 2,00 2.35 Nr wt MC 28d 1. 32 142. 1 5030 2115 2196
6 STRT 2,03 2.41 Nr wt MC 28d 1. 32 139.5 5030 2108 2184
8 STRT 2.30 2,55 Nr wt MC 28d 1.32 87.4 3760 1407 1365
9 STRT 2,33 2.41 Nr wt MC 28d 1. 32 90.0 3760 1461 1354
10 STRT 2.41 2.51 Nr wt MC 28d 1. 32 91.6 3760 1520 1416
a The eccentricities are measured values. bThese stresses refer to the steel cgs, and uses the measured values of F 0 and the net section
properties (+)compression;(-) tension. Remarks: All beams have a span= 19. 5'; all bars are 3/4" rp steel bars.; composite slabs (26" x 3") were cast on beams 1, 4, 6, 8 & 10 at 101, 101, 101, 37 & 93 days after stressing. The steel bars (of Es= 26380 ksi) were not grouted. Beams 1-8 were stored in the lab at 75% humidity & 9-10 were stored in the field at 90% R.H. The mix had a cement content of 6-6. 5 sacks/cu yd of Type I cement.
N
TABLE C2 PROPER TIES OF TEST BEAMS AT UNIVERSITY OF ILLINOIS (24)
c Eccentricitu a I
b Cone, Stress Cable Concrete Curing Loading A Fi fc rel Beam in2 Profile End Center Type Cond, Age (kips) (psi) End (psi) Center (pa i)
~U-1 STR 1. 03" 1. 03" Nr wt MC 5d . 18 26.9 3760 1285 1266
~U-2 STR 1.03" 1. 03" Nr wt MC 5d . 18 26.9 3930 1230 1271
a All eccentricities are measured values.
b These stresses refer to the steel cgs and uses the stress diagram indicated in Reference (24) (+)Compression;(-) Tension.
c All beams were cast of Type III cement with a water-cement ratio of 0. 74-0, 76 and a ratio of (1:2. 98:3, 35) of cement, sand and gravel by wt.
Remarks: All beams have a span= 6 1 , all wires are .196" cp (Es= 30 x 103 ksi). All beams were stored at 50% RH.
!l> 'tl N N
TABLE C3 PROPERTIES OF TEST BEAMS AT TEXAS A & M UNIVERSITY(~)
a b
Cone. Stress As Fi
I
Beam Length Span Cable Eccentricity Cone. Curing Load. fc rel End Center
Profile End Center Type Cond. Age in2 (k) psi {psi) (psi)
a All concrete stresses are computed using Fi and the transferred section properties. {+ compression), (- tens ion) and are at the steel cgs.
b All girders have the section designated as Type B by the Texas Highway Department.
Remarks: All strands are 7/16 11 <p (E8
= 28500 ksi) at an average humidity of 88%. The mix had a cement content of 7 - 7-1/2 s c/cu yd of Type III cement. All harping was at 5 1 from 1:_ of the girder.
TABLE C4 PROPERTIES OF TEST BEAMS AT UNIVERSITY OF MISSOURI (1l_)
Beam Cable Eccentricitu Concrete Curing Load As bF., £~ rel Cone. Stress Profile End Center Type Cond. Age in2 k psi End (psi) Center (psi)
East Para- . 83" 27.55" Nr wt MC 6ld 2.65 452.4 5160 682 1440 Girder bolic
a West Para- . 83" 27.55" Nr wt MC 37d 2.65 450.3 5190 - - - -Girder bolic
a Data from West Girder not available in this reference.
b This force F 0 is after el. losses and is the measured value of the force at the end. The value of F
0 at the center has been estimated from the strain measurements. The steel had an
(Es = 28. 8 x 10 3 ksi).
Remarks: Both girders had a span= 88 1; slab cast at age of concrete of 146d, 54 no of 1/4" cp
strands; 3 diaphrams at 24 1 -10", 49'-6", 74'-2" from end; these are shared at 1/4, 1/2, 3/4 points and stored at 70%.
Ap 25
TABLE CS
aDETAILS OF BEAMS REPORTED BY ABELES(~)
Area b Eff. Ecc. of cModulus Meas. of pres t. prest. of rupture cone.
steel force Ft' steel I
strength Type of fcb As (in2) (in) (psi)
I . Beam cone. (kips) fc (psi)
A01':' Nr wt . 2848 37.2 1, 25 570 5725
ALl'~ Lt wt .2848 32.5 1. 25 486 6600
aThe beams were 4 11 x 9" in section and simply supported on a span of 13' -9". A two point loading symmetrical about the center line of beam (i.e., at a distance of 5' fr om either support) was used for the test.
b The value of the effective pres tr es sing force is based on the reported magnitude of the effective pres tress.
c r.:t The modulus of rupture was based on a value of 6 ,.J fc concrete and 7. 5 fi'c for normal weight concrete.
for lightweight
Remarks: . 0712 in2 of steel area was provided as compressive reinforcement for both beams. The measured steel stress at ultimate was 240 ksi.
Beam
'° N .... • ~
"' µ:i p:;
"' °' 0 . ~
"' µ:i p:;
.... "' 0 . ~
"' µ:i p:;
Area of
Ap 26
TABLE C6
a DETAILS OF BEAMS REPORTED BY WARAWARUK, SOZEN & SIESS (Q)
Eff. Ecc. of Modulus Pre stress Pres tress of
steel, As force, Ft steel rupture (in2 ) (kips) (in)
I
fcb• psi
• 362 40.6 3.08 543
• 211 24. l 3.06 472
• 091 10.8 3,00 544
Meas. cone.
strength I
fc, psi
5230
3970
5280
a The beams were 6 11 x 12" in section and simply supported on a span of 9' -0". A two point loading symmetrical about the center line of the beam (i.e., at a distance of 3'-0" from either support) was used for the test.
TABLE C7
DETAILS OF BEAMS REPORTED BY SHAIKH AND BRANSON (49)
All beams 6" by 8", All d = 6. 5", All span= 15' simply supported
c I (in2 ) As .200 • 400 .600 . 058 • 200 • 400 0 0 . 310 • 080 • 310 .600
d, "'f c in psi 5400 5890 6570 5880
e Modulus of
806 855 I . 894 830 rupture, fcb(ps1)
a The value of the effective pres tress force, Ft was determined as Fi - ii Ft was determined using relationships developed in Reference (il).
b Refers to total tensile reinforcement (tensioned only).
c Refers to total non-tensioned tensile reinforcement.
d Refers to concrete strength at 28 days.
e Refers to modulus of rupture of concrete as measured from laboratory tests on plain concrete
specimens.
Ap 28
TABLE CB
a DETAILS OF BEAMS REPORTED BY BURNS & SIESS (54)
b c I d I e f As fcb fc Beam (in2 }
Eccentricity y Remarks (psi) (psi) (psi)
J9 1,58 8.00 510 4190 47.0 All beams had a
span of 12 1 -0 11;
JlO 1. 58 6.00 474 3590 45. 1 The reinforcing
steel had on elas -
Jll 1. 58 4.00 505 4110 46.9 ticity modulus of
30 x 106
psi.
a All beams had a width of 8". The total depth for beams J9, JlO, and Jll was 20", 16 11 and 12" respectively. All beams were centrally loaded.
b Refers to the total tensile reinforcement.
c Refers to the modulus of rupture of concrete at the time of the test.
d Refers to the concrete strength at the time of the test.
e Refers to the yield strength of the reinforcement.
Ap 29
APPENDIX D
Appendix D includes the details of the common
cases of prestress moment profiles along with
the formulas for computing camber.
Ap 30
APPENDIX D
COMMON CASES OF PRESTRESS MOMENT DIAGRAMS WITH FORMULAS FOR COMPUTING CAMBER
Prestressed Beam F 0 e Moment Diagram
Midspan Camber Due to F
0 e Moments
e l OT
(A1· )F = F eL2 /8 E .r O Cl g
0
12 E . I Cl g
(A ) _ F (e +e ) ~L2 Zj F e L2 i F - o c o _ -~ _ o o o E . I 8 6 SE .I
Cl g Cl g
Ap 31
APPENDIX E
Appendix E includes photographs of the laboratory
specimens during the various stages of the experi
mental program.
Ap 32
Figure El View of laboratory showing beams in foreground and prestressing bed containing additional beams at right.
Figure E2 Forms for beams in prestressing bed
Ap 33
Figure E3 Strain gage indicator and switching and balancing unit used with load cells to measure prestress force
Figure E4 Pres tressing bed, jacking equipment and beams stored in bed
Ap 34
Figure ES Close-up of jacking equipment, bulkheads, and grips
Figure E6 Shrinkage specimens in foreground and 7 beams ( 1 beam c ros swis e in foreground). Two additional beams in prestressing bed
Ap 35
Figure E7 Two of 4 composite beams. Strain gage points and dial gages can be seen. Strands used in relaxation tests are seen at right
Figure EB Cylinders loaded in creep racks and Whittemore gage used to measure strains of beams and shrinkage and creep specimens
Ap 36
Figure E9 View of beam Cl showing the crack pattern prior to failure
Figure E 10 View of beam C 1 after failure
Ap 37
APPENDIX F
Appendix F includes the following:
(i) A 'loss of prestress and camber' flow chart, its explanation and a typical computer output for interior girder No, 153.
(ii) A 'load-deflection' flow chart, its explanation and a typical computer output for laboratory beam Al.
Ap 38
Start
Read Input Data
{For details, see explanation of flow chart)
Write Input Data
Initialize all variables
Compute the correct ultimate creep and shrinkage coefs using
a sub•routine
Compute the elasticity modulii of beam con-crete at release and at slab casting as well as the elasticity modulus of slab concrete at
28 days
6
Ap 39
Compute the moments and the deflections (initial} due to slab and diaphram loads
Compute the initial loss of pres tress at end and center of the
beam
Determine the effective initial prestress force (after elastic
losses}
Compute the shrinkage and creep coefficients for 'time' required as well as for the ultimate conditions based on type of curing of the beam and slab concrete. Depending on the 'time' parameter, determine the loss of prestress due to shrinkage.
Ap 40
Depending on the 'time' parameter, compute the loss of prestress due to steel relaxation and due to creep of
concrete
Compute the total loss of prestres s
Compute the net initial camber due to prestress and beam dead
load
Compute the time -dependent camber due to prestress and time-dependent deflection due to beam dead load. Compute the total camber if 'time 1 parameter corresponds to a period prior to
slab casting
3
Ap 41
Write results if 1time 1 par Continue if 'time' parameter ameter corresponds corresponds to a period after 1------lperiod prior to
slab casting slab
End
C~mpute the loss of pres tress at end and center due .....- !( •
to creep depending on the type of curing of the beam
concrete. Similar values are obtained corresponding
to the 'ultimate' stage also. Compute the elastic and
creep gains due to the slab and diaphram loadings.
Compute the gains due to differential shrinkage also.
Ap 42
Compute the total loss of prestress corresponding to the 'time' parameter and the 'ulti-
mage' stage
Compute the initial camber due to prestress, the initial
deflection due to beam dead load and time -dependent camber
and deflection prior to and after slab casting due to prestress
and beam dead load respectively. Also compute the time-
dependent deflection due to slab plus diaphram loading.
Determine the deflection due to differential shrinkage and
then the total camber corresponding to the 'time' parameter
and the 'ultimate' stage.
Write results of analysis
Ap 43
EXPLANATION OF FLOW CHART FOR LOSS AND CAMBER
SL No.
1-2 1
22-50
51-96
97-98
99-115
116-120
121-134
135-139
140-190
Explanation
The read-in data includes the unit weight of beam concrete, unit weight of slab concrete, beam concrete strength at release, beam concrete strength at 28 days, slump of beam concrete, slab concrete strength at 28 days, ultimate shrinkage coefficient of slab concrete, elasticity modulus of prestressing steel, gross properties of the beam section, initial prestressing force, ultimate creep and shrinkage coefficients of beam concrete (referred to standard conditions), thickness and gross area of slab section,relative humidity, age of beam concrete at release of prestress and at slab casting, identifiers for type of curing and type of cement for beam concrete, diaphram loading and diaphram deflection, composite section properties, time parameter, and the correction factors for creep and shrinkage coefficients for the 'ultimate' stage.
Write input data.
Initialize all variables.
Compute the correct ultimate creep and shrinkage coefficients using a sub-routine.
Compute the elasticity modulii of beam concrete at release and at slab casting as well as the elasticity modulus of slab concrete at 28 days.
Compute the moments and the deflections (initial) due to slab and diaphram loads.
Compute the initial loss of prestress at end and center of the beam.
Determine the effective initial prestress force (after elastic loss es).
Compute the shrinkage and creep coefficients for 'time' required as well as for the ultimate conditions based on type of curing of the beam and slab concrete. Depending
191-209
210-213
214-225
226-237
?38-307
308-352
353-386
389-467
Ap 44
on the 'time' parameter, determine the loss of prestress due to shrinkage.
Depending on the 'time' parameter, compute the loss of prestress due to steel relaxation and due to creep of concrete.
Compute the total loss of prestress.
Compute the net initial camber due to prestress and beam dead load.·
Compute the time-dependent camber due to prestress and time -dependent deflection due to beam dead load. Compute the total camber if 'time' parameter corresponds to a period prior to slab casting.
Compute the loss of prestress at end and center due to creep depending on the type of curing of the beam concrete. Similar values are obtained corresponding to the 'ultimate' stage also. Compute the elastic and creep gains due to the slab and dis phram loadings. Compute the gains due to differential shrinkage also.
Compute the initial camber due to pres tress, the initial deflection due to beam dead load and time -dependent camber and deflection prior to and after slab casting due to prestress and beam dead load respectively. Also compute the time-dependent deflection due to slab plus diaphram loading. Determine the deflection due to differential shrinkage and then the total camber corresponding to the 'time' parameter and the 'ultimate' stage.
Write results of analysis.
This is a sub-routine to apply correction factors for the ultimate values of creep and shrinkage coefficients.
Ap 45
SJ(l~ 1 KRJPf\ 1
1 cnKMO~:/rN.:: /CU, 2SHU,(ll1" ,cus ,(lJ,.'5 ,cu~ s, ESt--U5 'TF. SHU, T t 1 Ts' TK l ,H, ~ESHU~,ru.s!,CGQ4,Srk4,CrRTT,CC~T
24 hRllf 16,33333) 25 33~33 FCRf'fll(lll ,!4X,'l NP UT 0 AT A '//1 26 •RlTElc,22001 27 2200 f(':1:1"~Tfl~! ,10.x, 1 !3 ~AP CC f\ C PETE'///} 28 hRIT~(t,20l)i:l,FCR,rCl2B 1 Sl 29 201 FCl-lf'AT(ll1 ,::sx,•uNtT WT. (PCF) •F25.2.//
*4X, 1 CC\C. 5Ti.; .• r,r t~EL. CPSll 1 fl4.2// ~4X 1 '(Ct\C. STR. AT 2.~-DAY IPSI) 1 r14.?// *4X, 1 C(t,(. SLU/1» I I!·~ChESJ 1 Fl4.2//)
30 hR I H' I 6 , .? 2 0 l I 31 ,20! FCc~>Tf lH ,1ox.•s l A B cc~ c p E TE '///I 32 ~RIT~tt,?03)W2,FCZLB. 33 203 FOPf'IO.Tt lit ,3X, 1 U\Jl T ~JT (P(F) 1 F25.2//
*4X, •cc:·.c. ST~1'\. <"SI J •Fl4.2//) 34 WRIT:=t:'-,204) 35 204 FCP"1Tlll-I ,JO<,•~' AM SECT l C N '1111 36 ~RIT~l~12051~G 1 G[,2E,~C,SP,~S,TK1,~P,Fl1~Sl1CORT,CCRTT 37 205 F'O?V1TllH ,3:<,'GPCSS ,\P[J\ ll~'*'(t21 1 F25.2//
38 "'RfT;::(t.120hl 39 206 FCHfi·\l(i>--1 ,1 X'S L fl f: SECT IC N '///) 40 ;-1'<IT:::(1:,::-011rK2,rir;2,cL[;"'',(lEFCtJ 41 207 Fr.:::1V1iTI !rl ,3X, 1 SLAR THK. ( l~I •F75."Z//
-1<4X, 'Sl-1/! J,~_::!\ I 1·'J*4.tl •FJ4.;:>// *4X, 1 CltJ ou:· TO Dl•\µh. ct~J-KlPSJ 1 Fl4.('//
Ap 46
•4x,•r.Er. DUE fr) 01•. 11~1 'fl4.21/l 42 WRIT•to,~0•1 43 208 FCO,OTll!t .iox,'l I M f - 0 f p E ~ 0 E 'I 1 FA c T 0 R S'///I 44 hRIT~{~,/OGJT1,TS,H 45 20~ fOH,'11111 ,3x,•r.~ ACE OT REL (CAYl'F24.2//
•4X 0 '1~' OF 6~ AT SLAH Cl5Tt01Y5l •fl4.2// •4x,••v·R~GE ~EL. H~'· (0/01 'Fl4.Z//I
46 oRITE(6,4001 47 400 rc~1·•T<1t< .iox,•c r' P • s E c 1 • c ET A I l s•11n 49 W~IF<•·'•Ol lCI 1,vccs.scc,•cc.R.HIC 49 401 FOP.Wf.T(lll ,3x,•COMP. "I ·(J"-!••4J 1 f25.2//
UL 1 f\ ! = -Qt_I LT *YC c s:i:•s p,,, Si>* l '· 4 ·' ( e • * E cs O:<C 1 I ) tlLT.b2=T!::;..;~~! ULTSR•l-IGFlULT-RAVGll+ll.-IDFlULT•RAVGlJ00.5ll*ICCSP-CTTTI
UL f_A4=T::r>f'J51 ULT1"•-ICCPP-CTTTl•CELTl20RIT!C UL r A6=H<nl LLTA7=-CUSS•nELL*RATIC*CCRT/CCR4 ULT AB=-CllMS•QoLL*R>T I C<•COU /CCR4 ULfA9=-<CCSP-CTTT)¥CELTA2*~hTIC ~RIT~(0,2101c2:,c3c,FO,DELTA
OLM CLf TC GI AP 1-. I IN-KIPS) 776 .oo DEF. CLE TC c j,I. I l 'll 0.23
T I r. E: - 0 s p E N 0 F. N T F A c T c R s
o~ AGE >T REL lC'YI 2.00
A Ge CF "" hT Slflt1 c~srw~vs1 67.00
AV!:fL~GC REL. HL~. ( """) 10.00
c 0 N p . s !? c T . D E T A l L s
CO•P. rt Clf\**4) 331167.00
F:CC. CF DI FF. F CP.C E t I': I 13. 56
CGS GIST AT !:NC I I~ I 29.'20
CGS CIST >T CTR I I« I 21.20
RAT IC IUIC o.33
C 0 r P L T E D ~ E S U L T S
CREEP •~D SHRl~KIGE CCEFFS l~CLUOl~G CDRR FICTCRS
ULT.CRP CCEFF--•,C,
ULT.C~P.CCEfF--s.c.
ULT. CPP CCEFF-- SLL C• S.C.
ULT. c~· CCEFF-- SL• c~ r.c.
lfLT.S~~K cc:FF--~k2CjSJ fi1~
ULT. S~RK CC~FF--SLI'. FkC~ O~Yl
o.oo 1.62
o.oo 352,,EO
330.00
Ap 56
N I T I A L S T A T E
EL. LCSS lE~DI 9.01
FL. LCSS lCTRI 12.03
PMES. FrRCE PO lKIPSI
INITIAL CAPeER llNI
LCSS AT BEA~ ENC 'T Tl~[ =
EL. LOSS
SHRK LCSS
CREEP eEFORE SLAO CAST
CREEP •FTER SL~C CAST
STEEL RELAX.
El• GAii:
CRoEP GAIN
GAIN CLE TC CIFF S~RINK
TOTAL LCSS
LCSS AT BEAP ENn AT ~LTl~ATE
FL. LCSS
St<RK LC SS
CRE~P efFCRf SLA3 CAST
CREEP ~fTER SLAP. C>ST
STEH R[LAX,
EL. GA!~
CRE<oP GAIN
GA!~ OLE TC ClfF s~rJ~K
TCHL LCSS
9.01
4. 54
7.69
i.oe
6.1~
o •. oo o.oo
-0.47
28.05
9.01
4.78
7.f~
1.10
7.!:0
o.oo o.oo
-0.44
30.24
5l0~0 CAYS
(L, LCSS
S11RK LCSS
CREEP e~FCRE SL•e C•ST
CREiP AFTER SLAE COST
HEEL HLAX.
EL. GAi~
CREEP GAIN
GAi~ CLF TC CJFF s~~r~·
TOTAL LCSS
Ap 57
12.03
4.2~
10.26
J.44
6.1>
-4.20
-l.3E
-0.64
LCSS AT BEAP CTR, AT ULTIMATE
fl• LcSS
SHRK LCSS
CREEP !EFCRE SLAB CAST
CREEP AFTER SLAE CtST
STEEL f!ELAX,
EL. G.01~1
CREEP G/.H:
GAIN CLE TG C!FF S~Al•K
TCT•L LCSS
P!DSP•N CA~BER ~T T [ t1 t
CBR c IJ ~ TC pq::s.
AP. CEA!'. LCtC CEFL.
CRP., Ct--P.~ QEFCRt $LB CAST
CRP c fJ i' P.. •FTER SL .:it\ C t..S T
CRP C EFL O'OFCRE SL 'fl CAST
CRP CEFL AFT[ R SL A·~ C~ST
EL. SLH' DEFL
CRP CtFL. cu~ TC SL 1\1)
560.0
12.0;
". 5 2
10.26
2.21
7.50
-4.20
-1.68
-0.Gl
30.03
CAYS
3. P.7
-l.64
2.3>
0.46
-1. 4 <;
-0.24
-2.21
-o. 73
Ap 58
DEFL, cu; TO C!FF, $11~)( -0.20
TCT6L CEFLC:CT!C~ O• cws:R 0.21
~ICSPAN CA HEH •T ~LTIMATE
CBR c~~ TC P~E·s • 3.87
ll~. CEJ.C LCAC C~Fl. -1.64
CRP. C~P.~ ~EFCR( SL£ CAST 2,3q
CRP c•eR. AFTER SL/oC CAST o. 71
CRP CEFL HFCRt SLAB CAST -1.~q
CRP CEFL AFT::R SLAe CAST -0.38
EL, Sl~B CEFL -2.21
CRP CE fl, cu~ TC SLAB -o.eq
DEFL, cu~ TO CJ FF, SH~K -o.1q
TOTAL CEFLECT!C~ CR CH'e"R 0 •. 11
Ap 59
FLOW CHART FOR LOAD-DEFLECTION STUDIES
Read input data (for details, see explanation
of flow chart)
Write Input Data
DO Cl I = l, 70
IF (Load. GT. Ultimate Load)
Compute effective moment of inertia
and then the deflection under applied
load
I
I I I I I I I I I I
"'
Ap 60
Write Results of Analysis
r I r I I
-- - ------ -- ___ .J
Ap 61
EXPLANATION OF FLOW CHART FOR LOAD DEFLECTION STUDIES
SL No.
1-5
5-7
8-13
14-31
Explanation
The read-in data includes the beam dead load, effective prestress force at the time of test, concrete modulus of rupture, gross sectional properties of the beam, the concrete strength at the time of test, the ultimate load of the beam, the cracking load of the beam and the cracked moment of inertia.
Compute the maximum dead load moment of the beam and the cracking moment of the beam.
Write pertinent information from the read-in data.
Compute the deflection under applied load and print the results.